<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD Journal Publishing DTD v2.3 20070202//EN" "journalpublishing.dtd">
<article article-type="research-article" dtd-version="2.3" xml:lang="EN" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">
<front>
<journal-meta>
<journal-id journal-id-type="publisher-id">Front. Bioinform.</journal-id>
<journal-title>Frontiers in Bioinformatics</journal-title>
<abbrev-journal-title abbrev-type="pubmed">Front. Bioinform.</abbrev-journal-title>
<issn pub-type="epub">2673-7647</issn>
<publisher>
<publisher-name>Frontiers Media S.A.</publisher-name>
</publisher>
</journal-meta>
<article-meta>
<article-id pub-id-type="publisher-id">716112</article-id>
<article-id pub-id-type="doi">10.3389/fbinf.2021.716112</article-id>
<article-categories>
<subj-group subj-group-type="heading">
<subject>Bioinformatics</subject>
<subj-group>
<subject>Original Research</subject>
</subj-group>
</subj-group>
</article-categories>
<title-group>
<article-title>Dynamic Solution Space Division-Based Methods for Calculating Reaction Deletion Strategies for Constraint-Based Metabolic Networks for Substance Production: DynCubeProd</article-title>
<alt-title alt-title-type="left-running-head">Ma and Tamura</alt-title>
<alt-title alt-title-type="right-running-head">DynCubeProd</alt-title>
</title-group>
<contrib-group>
<contrib contrib-type="author">
<name>
<surname>Ma</surname>
<given-names>Yier</given-names>
</name>
</contrib>
<contrib contrib-type="author" corresp="yes">
<name>
<surname>Tamura</surname>
<given-names>Takeyuki</given-names>
</name>
<xref ref-type="corresp" rid="c001">&#x2a;</xref>
<uri xlink:href="https://loop.frontiersin.org/people/799258/overview"/>
</contrib>
</contrib-group>
<aff>Bioinformatics Center, Institute for Chemical Research, Kyoto University, <addr-line>Kyoto</addr-line>, <country>Japan</country>
</aff>
<author-notes>
<fn fn-type="edited-by">
<p>
<bold>Edited by:</bold> <ext-link ext-link-type="uri" xlink:href="https://loop.frontiersin.org/people/142613/overview">Hongmin Cai</ext-link>, South China University of Technology, China</p>
</fn>
<fn fn-type="edited-by">
<p>
<bold>Reviewed by:</bold> <ext-link ext-link-type="uri" xlink:href="https://loop.frontiersin.org/people/809358/overview">Yushan Qiu</ext-link>, Shenzhen University, China</p>
<p>
<ext-link ext-link-type="uri" xlink:href="https://loop.frontiersin.org/people/1381345/overview">Xiaoqing Cheng</ext-link>, Xi&#x2019;an Jiaotong University, China</p>
</fn>
<corresp id="c001">&#x2a;Correspondence: Takeyuki Tamura, <email>tamura@kuicr.kyoto-u.ac.jp</email>
</corresp>
<fn fn-type="other">
<p>This article was submitted to Network Bioinformatics, a section of the journal Frontiers in Bioinformatics</p>
</fn>
</author-notes>
<pub-date pub-type="epub">
<day>13</day>
<month>08</month>
<year>2021</year>
</pub-date>
<pub-date pub-type="collection">
<year>2021</year>
</pub-date>
<volume>1</volume>
<elocation-id>716112</elocation-id>
<history>
<date date-type="received">
<day>28</day>
<month>05</month>
<year>2021</year>
</date>
<date date-type="accepted">
<day>13</day>
<month>07</month>
<year>2021</year>
</date>
</history>
<permissions>
<copyright-statement>Copyright &#xa9; 2021 Ma and Tamura.</copyright-statement>
<copyright-year>2021</copyright-year>
<copyright-holder>Ma and Tamura</copyright-holder>
<license xlink:href="http://creativecommons.org/licenses/by/4.0/">
<p>This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these&#x20;terms.</p>
</license>
</permissions>
<abstract>
<p>Flux balance analysis (FBA) is a crucial method to analyze large-scale constraint-based metabolic networks and computing design strategies for strain production in metabolic engineering. However, as it is often non-straightforward to obtain such design strategies to produce valuable metabolites, many tools have been proposed based on FBA. Among them, GridProd, which divides the solution space into small squares by focusing on the cell growth rate and the target metabolite production rate to efficiently find the reaction deletion strategies, was extended to CubeProd, which divides the solution space into small cubes. However, as GridProd and CubeProd naively divide the solution space into equal sizes, even places where solutions are unlikely to exist are examined. To address this issue, we introduce dynamic solution space division methods based on CubeProd for faster computing by avoiding searching in places where the solutions do not exist. We applied the proposed method DynCubeProd to iJO1366, which is a genome-scale constraint-based model of <italic>Escherichia coli</italic>. Compared with CubeProd, DynCubeProd significantly accelerated the calculation of the reaction deletion strategy for each target metabolite production. In addition, under the anaerobic condition of iJO1366, DynCubeProd could obtain the reaction deletion strategies for almost 40% of the target metabolites that the elementary flux vector-based method, which is one of the most effective methods in existence, could not. The developed software is available on <ext-link ext-link-type="uri" xlink:href="https://github.com/Ma-Yier/DynCubeProd">https://github.com/Ma-Yier/DynCubeProd</ext-link>.</p>
</abstract>
<kwd-group>
<kwd>metabolic network</kwd>
<kwd>flux balance analysis</kwd>
<kwd>constraint-based model</kwd>
<kwd>linear programming</kwd>
<kwd>algorithm</kwd>
</kwd-group>
</article-meta>
</front>
<body>
<sec id="s1">
<title>1 Introduction</title>
<p>Metabolic engineering is a DNA recombination-based technology proposed in 1991 to improve the designated substance production and the cell properties by manipulating and introducing specific biochemical reactions (<xref ref-type="bibr" rid="B2">Bailey, 1991</xref>; <xref ref-type="bibr" rid="B13">Stephanopoulos et&#x20;al., 1998</xref>). In many cases, current metabolic engineering technology focuses on the utilization of microorganisms. In metabolic engineering analysis, metabolic pathways in organisms are often represented by metabolic networks, in which nodes represent metabolite molecules and biochemical reactions. Any two metabolites (biochemical reactions) cannot be directly connected, and a metabolite must be connected to at least two biochemical reactions. The biochemical reactions can be irreversible or reversible. Nodes of external metabolites form the input and output nodes of the entire network.</p>
<p>Constraint-based modeling is a mathematical method to identify the best solution within a set of possible choices subject to pre-specified constraints (<xref ref-type="bibr" rid="B8">Maranas and Zomorrodi, 2016</xref>). Constraint-based modeling methods, such as linear programming (LP) and mixed integer linear programming, are widely used effective optimization techniques. Flux balance analysis (FBA) is one such widely used constraint-based modeling method with stoichiometric-based modeling of metabolism for the analysis of genome-scale metabolic models (GSMM) (<xref ref-type="bibr" rid="B8">Maranas and Zomorrodi, 2016</xref>).</p>
<p>In the constraint-based models of metabolic networks, the cell growth reaction and the target metabolite production reactions are of particular interest. The cell growth reaction has been virtually designed to simulate the efficient conversion of uptake resources into cellular energy and chemical components, which support cell growth in response to selection pressure to construct the system in the most plausible physiological state (<xref ref-type="bibr" rid="B8">Maranas and Zomorrodi, 2016</xref>). The target metabolite production reaction produces a chemical of interest. We define growth rate (GR) as cell growth reaction speed and production rate (PR) as the target metabolite production reaction&#x20;speed.</p>
<p>Growth coupling is a fundamental design principle in metabolic engineering and computational strain design. The purpose of growth coupling is to make the target metabolite a mandatory by-product of the cell growth reaction. We say that growth coupling is achieved if the target metabolite is produced when cell growth is maximized as shown in <xref ref-type="fig" rid="F1">Figure&#x20;1A</xref>.</p>
<fig id="F1" position="float">
<label>FIGURE 1</label>
<caption>
<p>
<bold>(A)</bold> In the problem setting of this study, it is required that the original network is converted into the designed network that achieves growth coupling, where cell growth and target metabolite production are simultaneously achieved. <bold>(B)</bold> The designed network should be obtained from the original network through reaction deletions. The black blocks are nutrient uptake reactions, the cell growth reaction, or the target metabolite production reaction. The speed of the cell growth reaction and target metabolite production reaction are represented by GR and PR, respectively. Light gray and dark gray represent the designed network and the original network, respectively.</p>
</caption>
<graphic xlink:href="fbinf-01-716112-g001.tif"/>
</fig>
<p>In this study, the core and basic problems are to find a growth coupling method for the target metabolite production through reaction deletions. The method should produce as much target metabolite as possible by modifying the metabolic network when GR is maximized as the objective function. The relationship between the original network and the designed network is shown in <xref ref-type="fig" rid="F1">Figure&#x20;1B</xref>. We can delete reactions by setting their speeds as zero in the network modification strategies. Based on the most basic problem above, the following sub-problems are derived. The first is to find knockout strategies for as many different target metabolites as possible. The second is to find knockout strategies for the networks under different input conditions such as aerobic or anaerobic conditions.</p>
<p>The most basic and pioneer algorithm for this purpose is OptKnock, which is a bilevel optimization-based method that identifies knockout strategies that result in the maximum PR when GR is maximized. The inner optimization performs the flux (reaction speed) allocation with regard to the optimization of cellular objectives (e.g., maximization of biomass yield and MOMA) (<xref ref-type="bibr" rid="B3">Burgard et&#x20;al., 2003</xref>). The outer optimization maximizes the bioengineering objective (e.g., chemical production) (<xref ref-type="bibr" rid="B3">Burgard et&#x20;al., 2003</xref>). However, because the computation time of OptKnock is proportional to an exponential function of the network size, in many cases, its computation is not completed within a realistic timeframe for GSMMs (<xref ref-type="bibr" rid="B16">Tamura, 2021b</xref>). Therefore, many algorithms have been proposed to speed up the process for the efficient computation of the reaction deletion strategies.</p>
<p>Considering that finding the optimal strategy is NP-hard, it is reasonable to only find out the strategy that meets the expected requirements. For example, the elementary flux vector (EFV)-based method determines reaction deletion strategies in which cell growth forces the production of the target metabolite, and the success ratio of this method was very high under both anaerobic and aerobic conditions for several microbial models (<xref ref-type="bibr" rid="B17">von Kamp and Klamt, 2017</xref>). GridProd efficiently computes the design of minimum metabolic networks by using bilevel optimization approach with picking two-dimensional limits and gridding the constraint space (<xref ref-type="bibr" rid="B14">Tamura, 2018</xref>). CubeProd divides the entire constraint space into small cubes and gave good results for GSMM with extreme constraints (e.g., anaerobic condition) (<xref ref-type="bibr" rid="B15">Tamura, 2021a</xref>). The EFV-based method, GridProd, and CubeProd enable the calculation of reaction deletion strategies for many target metabolites that cannot be calculated using the previously developed methods. However, for <italic>Escherichia coli</italic> under anaerobic conditions, the reaction deletion strategies could not be obtained for many target compounds. In particular, for GridProd and CubeProd, the bottleneck was the computing speed. Therefore, it was expected to extend GridProd or CubeProd to shorten the computation&#x20;time.</p>
<p>In this study, we developed DynCubeProd that improves the computation speed of CubeProd. DynCubeProd employs a dynamic strategy for the cube sizes to obtain the same results as CubeProd; however its computation speed is much faster. The reaction deletion strategies obtained by DynCubeProd also supplement those of the EFV-based method under certain conditions. Under anaerobic conditions, we obtained the reaction deletion strategies for close to 40% of the target metabolites for which the EFV-based method could not determine strategies.</p>
</sec>
<sec id="s2">
<title>2 Materials and Methods</title>
<sec id="s2-1">
<title>2.1 Problem Definition</title>
<p>The general formalization of constraint-based modeling is as follows (<xref ref-type="bibr" rid="B8">Maranas and Zomorrodi, 2016</xref>):</p>
<p>
<bold>minimize (or maximize):</bold>
</p>
<p>&#x2003;&#x2003;&#x2003;<inline-formula id="inf1">
<mml:math id="m1">
<mml:mi>f</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi mathvariant="bold-italic">x</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
<p>
<bold>subject&#x20;to:</bold>
</p>
<p>&#x2003;&#x2003;&#x2003;<inline-formula id="inf2">
<mml:math id="m2">
<mml:mi>h</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi mathvariant="bold-italic">x</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>0</mml:mn>
</mml:math>
</inline-formula>
</p>
<p>&#x2003;&#x2003;&#x2003;<inline-formula id="inf3">
<mml:math id="m3">
<mml:mi>g</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi mathvariant="bold-italic">x</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x2264;</mml:mo>
<mml:mn>0</mml:mn>
</mml:math>
</inline-formula>
</p>
<p>&#x2003;&#x2003;&#x2003;<inline-formula id="inf4">
<mml:math id="m4">
<mml:mi mathvariant="bold-italic">x</mml:mi>
<mml:mo>&#x2208;</mml:mo>
<mml:mi>S</mml:mi>
</mml:math>
</inline-formula>
</p>
<p>
<inline-formula id="inf5">
<mml:math id="m5">
<mml:mi mathvariant="bold-italic">x</mml:mi>
</mml:math>
</inline-formula> is an <inline-formula id="inf6">
<mml:math id="m6">
<mml:mi>n</mml:mi>
</mml:math>
</inline-formula>-dimensional variable. <inline-formula id="inf7">
<mml:math id="m7">
<mml:mi>f</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi mathvariant="bold-italic">x</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> is the objective function to minimize or maximize. <inline-formula id="inf8">
<mml:math id="m8">
<mml:mi>S</mml:mi>
</mml:math>
</inline-formula> is the set from which the variable vector <inline-formula id="inf9">
<mml:math id="m9">
<mml:mi mathvariant="bold-italic">x</mml:mi>
</mml:math>
</inline-formula> ranges. <inline-formula id="inf10">
<mml:math id="m10">
<mml:mi>h</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi mathvariant="bold-italic">x</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> and <inline-formula id="inf11">
<mml:math id="m11">
<mml:mi>g</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi mathvariant="bold-italic">x</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> are the constraints that must be satisfied as equalities or one-side inequalities, respectively.</p>
<p>The general form of the FBA is as follow:</p>
<p>
<bold>maximize</bold>
</p>
<p>&#x2003;&#x2003;&#x2003;<inline-formula id="inf12">
<mml:math id="m12">
<mml:mi>f</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi mathvariant="bold-italic">x</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
<p>
<bold>subject to:</bold>
</p>
<p>&#x2003;&#x2003;&#x2003;<inline-formula id="inf13">
<mml:math id="m13">
<mml:mi>S</mml:mi>
<mml:mi mathvariant="bold-italic">x</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>0</mml:mn>
</mml:math>
</inline-formula>
</p>
<p>&#x2003;&#x2003;&#x2003;<inline-formula id="inf14">
<mml:math id="m14">
<mml:mi>L</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>&#x2264;</mml:mo>
<mml:mi mathvariant="bold-italic">x</mml:mi>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>U</mml:mi>
<mml:mi>B</mml:mi>
</mml:math>
</inline-formula>
</p>
<p>
<inline-formula id="inf15">
<mml:math id="m15">
<mml:mi mathvariant="bold-italic">x</mml:mi>
<mml:mo>&#x2208;</mml:mo>
<mml:msup>
<mml:mrow>
<mml:mi mathvariant="double-struck">R</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msup>
</mml:math>
</inline-formula> is an <inline-formula id="inf16">
<mml:math id="m16">
<mml:mi>n</mml:mi>
</mml:math>
</inline-formula>-dimensional variable. <inline-formula id="inf17">
<mml:math id="m17">
<mml:mi>f</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi mathvariant="bold-italic">x</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> is the objective function, which in many cases is GR. <inline-formula id="inf18">
<mml:math id="m18">
<mml:mi>S</mml:mi>
<mml:mo>&#x2208;</mml:mo>
<mml:msup>
<mml:mrow>
<mml:mi mathvariant="double-struck">R</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>m</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msup>
</mml:math>
</inline-formula> is the stoichiometric matrix corresponding to <inline-formula id="inf19">
<mml:math id="m19">
<mml:mi>m</mml:mi>
</mml:math>
</inline-formula> metabolites and <inline-formula id="inf20">
<mml:math id="m20">
<mml:mi>n</mml:mi>
</mml:math>
</inline-formula> reactions in the constraint-based models. <inline-formula id="inf21">
<mml:math id="m21">
<mml:mi>L</mml:mi>
<mml:mi>B</mml:mi>
</mml:math>
</inline-formula> and <inline-formula id="inf22">
<mml:math id="m22">
<mml:mi>U</mml:mi>
<mml:mi>B</mml:mi>
</mml:math>
</inline-formula> impose the lower and upper bounds of each <inline-formula id="inf23">
<mml:math id="m23">
<mml:mi>x</mml:mi>
<mml:mo>&#x2208;</mml:mo>
<mml:mi mathvariant="bold-italic">x</mml:mi>
</mml:math>
</inline-formula>. For example, a flux for irreversible reactions <inline-formula id="inf24">
<mml:math id="m24">
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>i</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula> is constrained as <inline-formula id="inf25">
<mml:math id="m25">
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>i</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mn>0</mml:mn>
</mml:math>
</inline-formula>.</p>
<p>Our goal was to find reaction deletion strategies for growth coupling of target metabolite production. Let <inline-formula id="inf26">
<mml:math id="m26">
<mml:mi>K</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">{</mml:mo>
<mml:mrow>
<mml:msub>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo stretchy="false">&#x7c;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2208;</mml:mo>
<mml:mi>V</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">}</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> be a set of reactions to be knocked out, where <inline-formula id="inf27">
<mml:math id="m27">
<mml:mi>V</mml:mi>
</mml:math>
</inline-formula> is a set of <inline-formula id="inf28">
<mml:math id="m28">
<mml:mi>n</mml:mi>
</mml:math>
</inline-formula> reactions. Then, the definition of the main problem of this study arises.</p>
<p>
<bold>Given</bold>
</p>
<p>
<inline-formula id="inf29">
<mml:math id="m29">
<mml:mo>&#x2003;&#x2003;</mml:mo>
<mml:mi>S</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>L</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>U</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>,</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>,</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>,</mml:mo>
<mml:msubsup>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>m</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msubsup>
<mml:mo>,</mml:mo>
<mml:msubsup>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>d</mml:mi>
</mml:mrow>
</mml:msubsup>
</mml:math>
</inline-formula>
</p>
<p>
<bold>Find</bold>
</p>
<p>
<inline-formula id="inf30">
<mml:math id="m30">
<mml:mo>&#x2003;&#x2003;</mml:mo>
<mml:mi>K</mml:mi>
</mml:math>
</inline-formula>
</p>
<p>&#x2003;&#x2003;<bold>such That</bold>
</p>
<p>
<inline-formula id="inf31">
<mml:math id="m31">
<mml:msub>
<mml:mrow>
<mml:mo>&#x2003;&#x2003;&#x2003;&#x2003;&#x2003;</mml:mo>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:msubsup>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>m</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msubsup>
</mml:math>
</inline-formula> and <inline-formula id="inf32">
<mml:math id="m32">
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:msubsup>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>d</mml:mi>
</mml:mrow>
</mml:msubsup>
</mml:math>
</inline-formula>
</p>
<p>&#x2003;&#x2003;<bold>maximize</bold>
</p>
<p>&#x2003;&#x2003;&#x2003;&#x2003;&#x2003;<inline-formula id="inf33">
<mml:math id="m33">
<mml:mi>f</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> (&#x3d;<inline-formula id="inf34">
<mml:math id="m34">
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula>)&#x2003;&#x2003;&#x2003;<bold>subject&#x20;to:</bold>
</p>
<p>&#x2003;&#x2003;&#x2003;&#x2003;&#x2003;<inline-formula id="inf35">
<mml:math id="m35">
<mml:mi>S</mml:mi>
<mml:mi mathvariant="bold-italic">x</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>0</mml:mn>
</mml:math>
</inline-formula>
</p>
<p>&#x2003;&#x2003;&#x2003;&#x2003;&#x2003;<inline-formula id="inf36">
<mml:math id="m36">
<mml:mfenced open="{" close="">
<mml:mrow>
<mml:mtable class="cases">
<mml:mtr>
<mml:mtd columnalign="left">
<mml:mi>x</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>0</mml:mn>
<mml:mspace width="0.3333em" class="nbsp"/>
<mml:mi mathvariant="normal">i</mml:mi>
<mml:mi mathvariant="normal">f</mml:mi>
<mml:mspace width="0.3333em" class="nbsp"/>
<mml:mo>,</mml:mo>
<mml:mspace width="0.3333em"/>
<mml:mi>x</mml:mi>
<mml:mo>&#x2208;</mml:mo>
<mml:mi>K</mml:mi>
<mml:mspace width="1em"/>
</mml:mtd>
</mml:mtr>
<mml:mtr>
<mml:mtd columnalign="left">
<mml:mi>L</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>x</mml:mi>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>U</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>,</mml:mo>
<mml:mspace width="0.3333em"/>
<mml:mspace width="0.3333em" class="nbsp"/>
<mml:mi mathvariant="normal">o</mml:mi>
<mml:mi mathvariant="normal">t</mml:mi>
<mml:mi mathvariant="normal">h</mml:mi>
<mml:mi mathvariant="normal">e</mml:mi>
<mml:mi mathvariant="normal">r</mml:mi>
<mml:mi mathvariant="normal">w</mml:mi>
<mml:mi mathvariant="normal">i</mml:mi>
<mml:mi mathvariant="normal">s</mml:mi>
<mml:mi mathvariant="normal">e</mml:mi>
<mml:mo>.</mml:mo>
<mml:mspace width="1em"/>
</mml:mtd>
</mml:mtr>
</mml:mtable>
</mml:mrow>
</mml:mfenced>
</mml:math>
</inline-formula>
</p>
<p>When <inline-formula id="inf37">
<mml:math id="m37">
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:msubsup>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>m</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msubsup>
</mml:math>
</inline-formula> and <inline-formula id="inf38">
<mml:math id="m38">
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:msubsup>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>d</mml:mi>
</mml:mrow>
</mml:msubsup>
</mml:math>
</inline-formula> is satisfied, we consider <inline-formula id="inf39">
<mml:math id="m39">
<mml:mi>K</mml:mi>
</mml:math>
</inline-formula> achieves growth coupling, where <inline-formula id="inf40">
<mml:math id="m40">
<mml:mi>G</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula> for <inline-formula id="inf41">
<mml:math id="m41">
<mml:msub>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2208;</mml:mo>
<mml:mi>V</mml:mi>
</mml:math>
</inline-formula> and <inline-formula id="inf42">
<mml:math id="m42">
<mml:mi>P</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula> for <inline-formula id="inf43">
<mml:math id="m43">
<mml:msub>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2208;</mml:mo>
<mml:mi>V</mml:mi>
</mml:math>
</inline-formula>&#x20;hold.</p>
</sec>
<sec id="s2-2">
<title>2.2 Example for Problem Definition</title>
<p>A toy example of the constraint-based model with 11 nodes is shown in <xref ref-type="fig" rid="F2">Figure&#x20;2A</xref> to illustrate the problem definition explained above. The rectangular nodes <inline-formula id="inf44">
<mml:math id="m44">
<mml:mrow>
<mml:mo stretchy="false">{</mml:mo>
<mml:mrow>
<mml:mi>R</mml:mi>
<mml:mn>1</mml:mn>
<mml:mo>,</mml:mo>
<mml:mi>R</mml:mi>
<mml:mn>2</mml:mn>
<mml:mo>,</mml:mo>
<mml:mo>&#x2026;</mml:mo>
<mml:mo>,</mml:mo>
<mml:mi>R</mml:mi>
<mml:mn>7</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">}</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> are chemical reactions. <inline-formula id="inf45">
<mml:math id="m45">
<mml:mi>R</mml:mi>
<mml:mn>7</mml:mn>
</mml:math>
</inline-formula> is the target metabolite production reaction and <inline-formula id="inf46">
<mml:math id="m46">
<mml:mi>R</mml:mi>
<mml:mn>6</mml:mn>
</mml:math>
</inline-formula> is the cell growth reaction. The substrates and products of the reactions are shown on the right side of <xref ref-type="fig" rid="F2">Figure&#x20;2A</xref>. The gray rectangular nodes are external reactions which play roles of input and output of the entire network and the white rectangular nodes are internal reactions, each of which connect at least two metabolite nodes with different directions. The intervals next to the rectangular nodes are the lower and upper bounds of reaction speeds. The circular nodes are internal metabolites that connect rectangular&#x20;nodes.</p>
<fig id="F2" position="float">
<label>FIGURE 2</label>
<caption>
<p>
<bold>(A)</bold> A toy example of the constraint-based models. Rectangular nodes R1 to R7 are reactions, and the attached intervals represent lower and upper bounds of their reaction speeds. R1 is the nutrient uptake reaction. R6 is the cell growth reaction. R7 is the target metabolite production reaction. Circular nodes C1 to C4 are internal metabolites. <bold>(B)</bold> Reaction deletion strategies and the resulting flux (reaction speed) distributions.</p>
</caption>
<graphic xlink:href="fbinf-01-716112-g002.tif"/>
</fig>
<p>Suppose that <inline-formula id="inf47">
<mml:math id="m47">
<mml:msubsup>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>m</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msubsup>
<mml:mo>&#x3d;</mml:mo>
<mml:msubsup>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>d</mml:mi>
</mml:mrow>
</mml:msubsup>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula> is given. When a reaction deletion strategy is given and <inline-formula id="inf48">
<mml:math id="m48">
<mml:mi>G</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mn>6</mml:mn>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula> is maximized, if <inline-formula id="inf49">
<mml:math id="m49">
<mml:mi>G</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#x2265;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula> and <inline-formula id="inf50">
<mml:math id="m50">
<mml:mi>P</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mn>7</mml:mn>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula> hold, then we consider that growth coupling is achieved. Because this example is very simple, such a reaction deletion strategy can be easily determined through brute force enumeration. Deleting <inline-formula id="inf51">
<mml:math id="m51">
<mml:mi>R</mml:mi>
<mml:mn>3</mml:mn>
</mml:math>
</inline-formula> is the optimal solution for this toy example.</p>
<p>
<xref ref-type="fig" rid="F2">Figure&#x20;2B</xref> shows the results of each knockout strategy applied to this toy example. Because deleting R4 is practically equivalent to deleting R2, deleting R4 is omitted in the table. When none of the reactions are deleted, that is, <inline-formula id="inf52">
<mml:math id="m52">
<mml:mi>K</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>&#x3d5;</mml:mi>
</mml:math>
</inline-formula>, GR &#x3d; 3 will be obtained but PR will be 0. When <inline-formula id="inf53">
<mml:math id="m53">
<mml:mi>K</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">{</mml:mo>
<mml:mrow>
<mml:mi>R</mml:mi>
<mml:mn>2</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">}</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> or <inline-formula id="inf54">
<mml:math id="m54">
<mml:mi>K</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">{</mml:mo>
<mml:mrow>
<mml:mi>R</mml:mi>
<mml:mn>5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">}</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>, the same result will be obtained. According to the definition of the problem above, such knockout strategies are not acceptable. When <inline-formula id="inf55">
<mml:math id="m55">
<mml:mi>K</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">{</mml:mo>
<mml:mrow>
<mml:mi>R</mml:mi>
<mml:mn>3</mml:mn>
<mml:mo>,</mml:mo>
<mml:mi>R</mml:mi>
<mml:mn>5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">}</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>, PR &#x3d; 3 is obtained, but GR will be 0. When <inline-formula id="inf56">
<mml:math id="m56">
<mml:mi>K</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">{</mml:mo>
<mml:mrow>
<mml:mi>R</mml:mi>
<mml:mn>3</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">}</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>, both <inline-formula id="inf57">
<mml:math id="m57">
<mml:mi>G</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
<mml:mo>&#x2265;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula> and <inline-formula id="inf58">
<mml:math id="m58">
<mml:mi>P</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>2</mml:mn>
<mml:mo>&#x2265;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula> are obtained. Therefore, <inline-formula id="inf59">
<mml:math id="m59">
<mml:mi>K</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">{</mml:mo>
<mml:mrow>
<mml:mi>R</mml:mi>
<mml:mn>3</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">}</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> is a feasible solution because it achieves growth coupling. However, such a brute force method cannot be applied to GSSMs owing to combinatorial explosion.</p>
</sec>
</sec>
<sec id="s3">
<title>3 Results</title>
<p>We developed an algorithm DynCubeProd for calculating reaction deletion strategies that achieve growth coupling of designated target metabolite production in constraint-based models of metabolic networks.</p>
<p>Because DynCubeProd is a method obtained by improving CubeProd (<xref ref-type="bibr" rid="B15">Tamura, 2021a</xref>), in this section, we provide an overview of CubeProd and then explain the difference between DynCubeProd and CubeProd. The algorithm behavior of DynCubeProd is also illustrated using examples. The relationship between DynCubeProd and other methods is discussed in <xref ref-type="sec" rid="s4">Section&#x20;4</xref>.</p>
<sec id="s3-1">
<title>3.1 DynCubeProd</title>
<p>Idea: CubeProd considers the three-dimensional solution space whose axes represent GR, PR and sum of absolute values of fluxes (SF). Let TMGR, TMPR, and TMSF be the theoretical maximum values of the above, respectively. Then the whole constraint space is a rectangle formed by [0, TMGR], [0, TMPR], and [0,TMSF]<xref ref-type="fn" rid="fn1">
<sup>1</sup>
</xref>. According to the designated value of <inline-formula id="inf60">
<mml:math id="m60">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula>, each of [0, TMGR], [0, TMPR], and [0,TMSF] are divided into <inline-formula id="inf61">
<mml:math id="m61">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> pieces. Therefore, finally, CubeProd considers <inline-formula id="inf62">
<mml:math id="m62">
<mml:msup>
<mml:mrow>
<mml:mi>P</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mn>3</mml:mn>
</mml:mrow>
</mml:msup>
</mml:math>
</inline-formula> constraint sub-spaces.</p>
<p>The value of <inline-formula id="inf63">
<mml:math id="m63">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> closely affects the trade-off between the ease of finding a solution and the computation time. The larger the value of P, the easier it is to find a solution, but the slower the computation time. Therefore, if CubeProd uses a certain value of <inline-formula id="inf64">
<mml:math id="m64">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> and cannot find a solution, a larger value of <inline-formula id="inf65">
<mml:math id="m65">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> should be applied, but the computation time will be more. However, it may be the case that some of the small solution spaces generated by a larger <inline-formula id="inf66">
<mml:math id="m66">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> are already proved by a smaller <inline-formula id="inf67">
<mml:math id="m67">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> to contain no solutions.</p>
<p>DynCubeProd starts with <italic>p</italic>&#x20;&#x3d; 1 and doubles <inline-formula id="inf68">
<mml:math id="m68">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> if no solution is found. When applying a larger P, it refers to the result of applying the smaller <inline-formula id="inf69">
<mml:math id="m69">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> and avoids searching for places where there is no solution.</p>
<p>Because the intervals on each of the three axes are equally subdivided into <inline-formula id="inf70">
<mml:math id="m70">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> sub-intervals, the entire constraint space is divided into <inline-formula id="inf71">
<mml:math id="m71">
<mml:msup>
<mml:mrow>
<mml:mi>P</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mn>3</mml:mn>
</mml:mrow>
</mml:msup>
</mml:math>
</inline-formula> sub-spaces, and<disp-formula id="equ1">
<mml:math id="m72">
<mml:mfrac>
<mml:mrow>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>G</mml:mi>
<mml:mi>R</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>P</mml:mi>
</mml:mrow>
</mml:mfrac>
<mml:mo>&#x2264;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2264;</mml:mo>
<mml:mfrac>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>G</mml:mi>
<mml:mi>R</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>P</mml:mi>
</mml:mrow>
</mml:mfrac>
<mml:mo>,</mml:mo>
</mml:math>
</disp-formula>
<disp-formula id="equ2">
<mml:math id="m73">
<mml:mfrac>
<mml:mrow>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>j</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>P</mml:mi>
<mml:mi>R</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>P</mml:mi>
</mml:mrow>
</mml:mfrac>
<mml:mo>&#x2264;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2264;</mml:mo>
<mml:mfrac>
<mml:mrow>
<mml:mi>j</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>P</mml:mi>
<mml:mi>R</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>P</mml:mi>
</mml:mrow>
</mml:mfrac>
<mml:mo>,</mml:mo>
</mml:math>
</disp-formula>
<disp-formula id="equ3">
<mml:math id="m74">
<mml:mfrac>
<mml:mrow>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>k</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>S</mml:mi>
<mml:mi>F</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>P</mml:mi>
</mml:mrow>
</mml:mfrac>
<mml:mo>&#x2264;</mml:mo>
<mml:mo>&#x2211;</mml:mo>
<mml:mo stretchy="false">&#x7c;</mml:mo>
<mml:mi>x</mml:mi>
<mml:mo stretchy="false">&#x7c;</mml:mo>
<mml:mo>&#x2264;</mml:mo>
<mml:mfrac>
<mml:mrow>
<mml:mi>k</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>S</mml:mi>
<mml:mi>F</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>P</mml:mi>
</mml:mrow>
</mml:mfrac>
</mml:math>
</disp-formula>are added as constraints and the sum of the absolute values of fluxes is minimized for every <inline-formula id="inf72">
<mml:math id="m75">
<mml:mn>1</mml:mn>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula>, where <inline-formula id="inf73">
<mml:math id="m76">
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:math>
</inline-formula> are integers.</p>
<p>In each of those <inline-formula id="inf74">
<mml:math id="m77">
<mml:msup>
<mml:mrow>
<mml:mi>P</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mn>3</mml:mn>
</mml:mrow>
</mml:msup>
</mml:math>
</inline-formula> sub-spaces, 1) LP is employed with the above three constraints, 2) if the LP is feasible, reactions whose flux is less than <inline-formula id="inf75">
<mml:math id="m78">
<mml:mn>1</mml:mn>
<mml:msup>
<mml:mrow>
<mml:mn>0</mml:mn>
</mml:mrow>
<mml:mrow>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>5</mml:mn>
</mml:mrow>
</mml:msup>
</mml:math>
</inline-formula> are collected as <inline-formula id="inf76">
<mml:math id="m79">
<mml:mi>K</mml:mi>
</mml:math>
</inline-formula>, 3)the minimum value of PR is calculated with deletions of <inline-formula id="inf77">
<mml:math id="m80">
<mml:mi>K</mml:mi>
</mml:math>
</inline-formula> under the condition that GR is maximized without the above three constraints, and 4) if GR and PR exceed the minimum required values, the output <inline-formula id="inf78">
<mml:math id="m81">
<mml:mi>K</mml:mi>
</mml:math>
</inline-formula> is considered as the solution.</p>
<p>The number of sub-spaces to be computed is <inline-formula id="inf79">
<mml:math id="m82">
<mml:msup>
<mml:mrow>
<mml:mi>P</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mn>3</mml:mn>
</mml:mrow>
</mml:msup>
</mml:math>
</inline-formula> and it increases dramatically as <inline-formula id="inf80">
<mml:math id="m83">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> increases, which will lead to a power-of-three increase in computation time. However, the larger the value of <inline-formula id="inf81">
<mml:math id="m84">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula>, the smaller is the range of the subspace and the easier it is to approach the point of the optimal solution or local optimal solution.</p>
<p>A dynamic strategy is adopted by DynCubeProd to save time. Starting with <inline-formula id="inf82">
<mml:math id="m85">
<mml:mi>i</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula> for <inline-formula id="inf83">
<mml:math id="m86">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:msup>
<mml:mrow>
<mml:mn>2</mml:mn>
</mml:mrow>
<mml:mrow>
<mml:mi>i</mml:mi>
</mml:mrow>
</mml:msup>
</mml:math>
</inline-formula>, DynCubeProd increases <inline-formula id="inf84">
<mml:math id="m87">
<mml:mi>i</mml:mi>
</mml:math>
</inline-formula> one by one, and stops once an acceptable knockout strategy is obtained. Suppose that <inline-formula id="inf85">
<mml:math id="m88">
<mml:mi>Q</mml:mi>
</mml:math>
</inline-formula> is a sub-space corresponding to <inline-formula id="inf86">
<mml:math id="m89">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>m</mml:mi>
</mml:math>
</inline-formula> and <inline-formula id="inf87">
<mml:math id="m90">
<mml:mrow>
<mml:mo stretchy="false">{</mml:mo>
<mml:mrow>
<mml:msub>
<mml:mrow>
<mml:mi>Q</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mn>1</mml:mn>
</mml:mrow>
</mml:msub>
<mml:mo>,</mml:mo>
<mml:mo>&#x2026;</mml:mo>
<mml:mo>,</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>Q</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>k</mml:mi>
</mml:mrow>
</mml:msub>
</mml:mrow>
<mml:mo stretchy="false">}</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> are sub-spaces of <inline-formula id="inf88">
<mml:math id="m91">
<mml:mi>Q</mml:mi>
</mml:math>
</inline-formula> for <inline-formula id="inf89">
<mml:math id="m92">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>n</mml:mi>
</mml:math>
</inline-formula> with <inline-formula id="inf90">
<mml:math id="m93">
<mml:mi>m</mml:mi>
<mml:mo>&#x3c;</mml:mo>
<mml:mi>n</mml:mi>
</mml:math>
</inline-formula>. If the candidate knockout strategy computed from <inline-formula id="inf91">
<mml:math id="m94">
<mml:mi>Q</mml:mi>
</mml:math>
</inline-formula> is not acceptable when <inline-formula id="inf92">
<mml:math id="m95">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>m</mml:mi>
</mml:math>
</inline-formula>, all the sub-spaces <inline-formula id="inf93">
<mml:math id="m96">
<mml:mrow>
<mml:mo stretchy="false">{</mml:mo>
<mml:mrow>
<mml:msub>
<mml:mrow>
<mml:mi>Q</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mn>1</mml:mn>
</mml:mrow>
</mml:msub>
<mml:mo>,</mml:mo>
<mml:mo>&#x2026;</mml:mo>
<mml:mo>,</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>Q</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>k</mml:mi>
</mml:mrow>
</mml:msub>
</mml:mrow>
<mml:mo stretchy="false">}</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> will be skipped during the calculation when <inline-formula id="inf94">
<mml:math id="m97">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>n</mml:mi>
</mml:math>
</inline-formula>.</p>
<p>
<statement content-type="example" id="uExample_1">
<label>
<bold>Example:</bold>
</label>
<p>
<xref ref-type="table" rid="T1">Table&#x20;1A</xref> represents the whole solution space by DynCubeProd with <inline-formula id="inf95">
<mml:math id="m98">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula> of the toy network of <xref ref-type="fig" rid="F2">Figure&#x20;2A</xref> whereas <xref ref-type="table" rid="T1">Table&#x20;1B</xref> represents all sub-spaces for <inline-formula id="inf96">
<mml:math id="m99">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>2</mml:mn>
</mml:math>
</inline-formula>. For each sub-space, LPs are performed twice; first, to calculate the possible candidate knockout strategy and second, to verify that the candidate is acceptable. The results of two LPs are represented for each (sub)space in <xref ref-type="table" rid="T1">Tables&#x20;1A,B</xref>.</p>
</statement>
</p>
<table-wrap id="T1" position="float">
<label>TABLE 1</label>
<caption>
<p>DynCubeProd applies CubeProd beginning with <italic>p</italic> &#x3d; 1 and increments one by one and quits when a desired reaction deletion strategy is obtained. <bold>(A)</bold> and <bold>(B)</bold> represent the results of DynCubeProd applied to the network of <xref ref-type="fig" rid="F2">Figure 2A</xref> with <italic>p</italic> &#x3d; 1 and <italic>p</italic> &#x3d; 2, respectively.</p>
</caption>
<table>
<thead valign="top">
<tr>
<th align="left"/>
<th colspan="10" align="center">(A) CubeProd (<italic>p</italic> &#x3d; 1)</th>
</tr>
</thead>
<tbody valign="top">
<tr>
<td align="left"/>
<td align="center">PR</td>
<td align="center">GR</td>
<td align="center">SF</td>
<td align="center">R1</td>
<td align="center">R2</td>
<td align="center">R3</td>
<td align="center">R4</td>
<td align="center">R5</td>
<td align="center">R6</td>
<td align="center">R7</td>
</tr>
<tr>
<td rowspan="2" align="left"/>
<td rowspan="2" align="center">
<inline-formula id="inf109">
<mml:math id="m112">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,3</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf110">
<mml:math id="m113">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,3</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf111">
<mml:math id="m114">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,26</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
<td align="center">0</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
</tr>
<tr>
<td align="char" char=".">3</td>
<td align="center">0</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
<td align="center">0</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
</tr>
<tr>
<td align="left"/>
<td colspan="10" align="center">
<bold>(B) CubeProd (<italic>p</italic> &#x3d; 2)</bold>
</td>
</tr>
<tr>
<td align="left">Sub-space ID</td>
<td align="center">PR</td>
<td align="center">GR</td>
<td align="center">SF</td>
<td align="center">R1</td>
<td align="center">R2</td>
<td align="center">R3</td>
<td align="center">R4</td>
<td align="center">R5</td>
<td align="center">R6</td>
<td align="center">R7</td>
</tr>
<tr>
<td rowspan="2" align="left">1</td>
<td rowspan="2" align="center">
<inline-formula id="inf112">
<mml:math id="m115">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,1.5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf113">
<mml:math id="m116">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,1.5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf114">
<mml:math id="m117">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,13</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">3</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
<td align="center">0</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
</tr>
<tr>
<td align="char" char=".">3</td>
<td align="center">0</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
<td align="center">0</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
</tr>
<tr>
<td align="left">2</td>
<td align="center">
<inline-formula id="inf115">
<mml:math id="m118">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,1.5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">
<inline-formula id="inf116">
<mml:math id="m119">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,1.5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">
<inline-formula id="inf117">
<mml:math id="m120">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>13,26</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
</tr>
<tr>
<td rowspan="2" align="left">3</td>
<td rowspan="2" align="center">
<inline-formula id="inf118">
<mml:math id="m121">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,1.5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf119">
<mml:math id="m122">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1.5</mml:mn>
<mml:mo>,</mml:mo>
<mml:mn>3</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf120">
<mml:math id="m123">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,13</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">3</td>
<td align="center">0</td>
<td align="center">3</td>
<td align="center">0</td>
<td align="center">0</td>
<td align="center">3</td>
<td align="center">0</td>
</tr>
<tr>
<td align="char" char=".">3</td>
<td align="center">0</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
<td align="center">0</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
</tr>
<tr>
<td align="left">4</td>
<td align="center">
<inline-formula id="inf121">
<mml:math id="m124">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,1.5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">
<inline-formula id="inf122">
<mml:math id="m125">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1.5</mml:mn>
<mml:mo>,</mml:mo>
<mml:mn>3</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">
<inline-formula id="inf123">
<mml:math id="m126">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>13,26</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
</tr>
<tr>
<td rowspan="2" align="left">5</td>
<td rowspan="2" align="center">
<inline-formula id="inf124">
<mml:math id="m127">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1.5</mml:mn>
<mml:mo>,</mml:mo>
<mml:mn>3</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf125">
<mml:math id="m128">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,1.5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf126">
<mml:math id="m129">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,13</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">3</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
<td align="center">0</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
</tr>
<tr>
<td align="char" char=".">3</td>
<td align="center">0</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
<td align="center">0</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
</tr>
<tr>
<td align="left">6</td>
<td align="center">
<inline-formula id="inf127">
<mml:math id="m130">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1.5</mml:mn>
<mml:mo>,</mml:mo>
<mml:mn>3</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">
<inline-formula id="inf128">
<mml:math id="m131">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,1.5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">
<inline-formula id="inf129">
<mml:math id="m132">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>13,26</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
</tr>
<tr>
<td rowspan="2" align="left">7</td>
<td rowspan="2" align="center">
<inline-formula id="inf130">
<mml:math id="m133">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1.5</mml:mn>
<mml:mo>,</mml:mo>
<mml:mn>3</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf131">
<mml:math id="m134">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1.5</mml:mn>
<mml:mo>,</mml:mo>
<mml:mn>3</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf132">
<mml:math id="m135">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0,13</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">3</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
<td align="center">0</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
</tr>
<tr>
<td align="char" char=".">3</td>
<td align="center">0</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
<td align="center">0</td>
<td align="char" char=".">3</td>
<td align="center">0</td>
</tr>
<tr>
<td align="left">8</td>
<td align="center">
<inline-formula id="inf133">
<mml:math id="m136">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1.5</mml:mn>
<mml:mo>,</mml:mo>
<mml:mn>3</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">
<inline-formula id="inf134">
<mml:math id="m137">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1.5</mml:mn>
<mml:mo>,</mml:mo>
<mml:mn>3</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">
<inline-formula id="inf135">
<mml:math id="m138">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>13,26</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
<td align="center">null</td>
</tr>
</tbody>
</table>
</table-wrap>
<p>As shown in <xref ref-type="table" rid="T1">Table&#x20;1A</xref>, when <inline-formula id="inf97">
<mml:math id="m100">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula>, the number of sub-spaces is one; this implies that the sub-space is equivalent to the entire original constraint space. As TMGR, TMPR, and TMSF are 3, 3, 13, respectively, Three constraints <inline-formula id="inf98">
<mml:math id="m101">
<mml:mn>0</mml:mn>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>G</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#x2264;</mml:mo>
<mml:mn>3</mml:mn>
</mml:math>
</inline-formula>, <inline-formula id="inf99">
<mml:math id="m102">
<mml:mn>0</mml:mn>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>P</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#x2264;</mml:mo>
<mml:mn>3</mml:mn>
</mml:math>
</inline-formula> and <inline-formula id="inf100">
<mml:math id="m103">
<mml:mn>0</mml:mn>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>S</mml:mi>
<mml:mi>F</mml:mi>
<mml:mo>&#x2264;</mml:mo>
<mml:mn>2</mml:mn>
<mml:mo>&#x22c5;</mml:mo>
<mml:mn>13</mml:mn>
</mml:math>
</inline-formula> are added and the sum of absolute values of fluxes is minimized in the first LP. Then R1 &#x3d; R3 &#x3d; R6 &#x3d; 3 and R2 &#x3d; R4 &#x3d; R5 &#x3d; R7 &#x3d; 0 are obtained. Therefore, <inline-formula id="inf101">
<mml:math id="m104">
<mml:mi>K</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">{</mml:mo>
<mml:mrow>
<mml:mi>R</mml:mi>
<mml:mn>2</mml:mn>
<mml:mo>,</mml:mo>
<mml:mi>R</mml:mi>
<mml:mn>4</mml:mn>
<mml:mo>,</mml:mo>
<mml:mi>R</mml:mi>
<mml:mn>5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">}</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> is obtained as a reaction deletion strategy candidate because the target metabolite production reaction R7 is not allowed to be deleted. <inline-formula id="inf102">
<mml:math id="m105">
<mml:mi>K</mml:mi>
</mml:math>
</inline-formula> is validated by the second LP, however, GR &#x3d; 3 and PR &#x3d; 0 are obtained. Therefore, the deletion strategy <inline-formula id="inf103">
<mml:math id="m106">
<mml:mi>K</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">{</mml:mo>
<mml:mrow>
<mml:mi>R</mml:mi>
<mml:mn>2</mml:mn>
<mml:mo>,</mml:mo>
<mml:mi>R</mml:mi>
<mml:mn>4</mml:mn>
<mml:mo>,</mml:mo>
<mml:mi>R</mml:mi>
<mml:mn>5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">}</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> is not acceptable because PR does satisfy the minimum required value <inline-formula id="inf104">
<mml:math id="m107">
<mml:msubsup>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>d</mml:mi>
</mml:mrow>
</mml:msubsup>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula>.</p>
<p>Therefore, DynCubeProd with <italic>p</italic>&#x20;&#x3d; 2 is applied. The results are summarized in <xref ref-type="table" rid="T1">Table&#x20;1B</xref>. Because the number of axes in DynCubeProd is three, eight sub-spaces are created when <inline-formula id="inf105">
<mml:math id="m108">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>2</mml:mn>
</mml:math>
</inline-formula>. For the sub-space IDs 2, 4, 6, and 8 in <xref ref-type="table" rid="T1">Table&#x20;1B</xref>, there is no feasible solution in the first LP. Therefore, the computing in&#x20;<inline-formula id="inf106">
<mml:math id="m109">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>4</mml:mn>
</mml:math>
</inline-formula> will be skipped for all sub-spaces of the sub-spaces whose&#x20;IDs are 2, 4, 6, and 8 in <inline-formula id="inf107">
<mml:math id="m110">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>2</mml:mn>
</mml:math>
</inline-formula>. This process of skipping computing the sub-spaces that do not contain solutions makes DynCubeProd computation faster than CubeProd. DynCubeProd will end the computing and return results either when <inline-formula id="inf108">
<mml:math id="m111">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> reaches the designated maximum value or when an acceptable solution is obtained.</p>
<p>
<statement content-type="lemma" id="Lemma_1">
<label>
<bold>Lemma 1:</bold>
</label>
<p>
<italic>For a positive integer</italic> <inline-formula id="inf136">
<mml:math id="m139">
<mml:mi>k</mml:mi>
</mml:math>
</inline-formula>
<italic>, the solution obtained by DynCubeProd for</italic> <inline-formula id="inf137">
<mml:math id="m140">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>2</mml:mn>
<mml:mi>k</mml:mi>
</mml:math>
</inline-formula> <italic>is also a solution for</italic> <inline-formula id="inf138">
<mml:math id="m141">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>k</mml:mi>
</mml:math>
</inline-formula>
<italic>, and the constraint sub-spaces skipped by DynCubeProd do not include a solution.</italic>
</p>
</statement>
</p>
<p>
<statement content-type="proof" id="uProof_1">
<label>
<bold>Proof:</bold>
</label>
<p>Define <inline-formula id="inf139">
<mml:math id="m142">
<mml:mi>S</mml:mi>
</mml:math>
</inline-formula> as the constraint space of an LP problem. Let <inline-formula id="inf140">
<mml:math id="m143">
<mml:msub>
<mml:mrow>
<mml:mi>S</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mn>1</mml:mn>
</mml:mrow>
</mml:msub>
<mml:mo>&#x222a;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>S</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mn>2</mml:mn>
</mml:mrow>
</mml:msub>
<mml:mo>&#x222a;</mml:mo>
<mml:mo>&#x22ef;</mml:mo>
<mml:mo>&#x222a;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>S</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>S</mml:mi>
</mml:math>
</inline-formula> hold. Suppose there exists a solution <inline-formula id="inf141">
<mml:math id="m144">
<mml:mi>x</mml:mi>
</mml:math>
</inline-formula> in the sub-space <inline-formula id="inf142">
<mml:math id="m145">
<mml:msub>
<mml:mrow>
<mml:mi>S</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>k</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>,</mml:mo>
<mml:mspace width="0.3333em"/>
<mml:mi>x</mml:mi>
<mml:mo>&#x2208;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>S</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>k</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula>. Then, <inline-formula id="inf143">
<mml:math id="m146">
<mml:mi>x</mml:mi>
</mml:math>
</inline-formula> must be in the space of <inline-formula id="inf144">
<mml:math id="m147">
<mml:mi>S</mml:mi>
<mml:mo>,</mml:mo>
<mml:mspace width="0.3333em"/>
<mml:mi>x</mml:mi>
<mml:mo>&#x2208;</mml:mo>
<mml:mi>S</mml:mi>
</mml:math>
</inline-formula>, because <inline-formula id="inf145">
<mml:math id="m148">
<mml:msub>
<mml:mrow>
<mml:mi>S</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>k</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2282;</mml:mo>
<mml:mi>S</mml:mi>
</mml:math>
</inline-formula>. If there is no solution <inline-formula id="inf146">
<mml:math id="m149">
<mml:mi>x</mml:mi>
</mml:math>
</inline-formula> in <inline-formula id="inf147">
<mml:math id="m150">
<mml:mi>S</mml:mi>
</mml:math>
</inline-formula>, that is, <inline-formula id="inf148">
<mml:math id="m151">
<mml:mi>x</mml:mi>
<mml:mo>&#x2209;</mml:mo>
<mml:mi>S</mml:mi>
</mml:math>
</inline-formula>, then such <inline-formula id="inf149">
<mml:math id="m152">
<mml:mi>x</mml:mi>
</mml:math>
</inline-formula> must not exist in any sub-space of&#x20;<inline-formula id="inf150">
<mml:math id="m153">
<mml:mi>S</mml:mi>
</mml:math>
</inline-formula>.</p>
</statement>
</p>
</sec>
<sec id="s3-2">
<title>3.2 Pseudo Code of DynCubeProd</title>
<p>The pseudo code of DynCubeProd is as follow:</p>
<p>
<statement content-type="algorithm" id="ALGORITHM_1">
<label>
<bold>Algorithm 1:</bold>
</label>
<p>DynCubeProd <list list-type="simple">
<list-item>
<p>
<inline-formula id="inf151">
<mml:math id="m154">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mi>K</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>P</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>G</mml:mi>
<mml:mi>R</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
</mml:math>
</inline-formula>DynCubeProd (<inline-formula id="inf152">
<mml:math id="m155">
<mml:mi>S</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>L</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>U</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>G</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>O</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>N</mml:mi>
<mml:mi>G</mml:mi>
<mml:mi>A</mml:mi>
<mml:mi>M</mml:mi>
</mml:math>
</inline-formula>,</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf153">
<mml:math id="m156">
<mml:msub>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>,</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>,</mml:mo>
<mml:msubsup>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>m</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msubsup>
<mml:mo>,</mml:mo>
<mml:msubsup>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>d</mml:mi>
</mml:mrow>
</mml:msubsup>
<mml:mo>,</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>n</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>m</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>x</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula>)</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf154">
<mml:math id="m157">
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>G</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi mathvariant="bold-italic">m</mml:mi>
<mml:mi mathvariant="bold-italic">a</mml:mi>
<mml:mi mathvariant="bold-italic">x</mml:mi>
<mml:mspace width="0.3333em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf155">
<mml:math id="m158">
<mml:mspace width="2em"/>
<mml:mi mathvariant="bold-italic">s</mml:mi>
<mml:mo>.</mml:mo>
<mml:mi mathvariant="bold-italic">t</mml:mi>
<mml:mo>.</mml:mo>
<mml:mspace width="0.3333em"/>
<mml:mo movablelimits="false" form="prefix">&#x2211;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>S</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>0</mml:mn>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf156">
<mml:math id="m159">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>L</mml:mi>
<mml:msub>
<mml:mrow>
<mml:mi>B</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2264;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>U</mml:mi>
<mml:msub>
<mml:mrow>
<mml:mi>B</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf157">
<mml:math id="m160">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>c</mml:mi>
<mml:mtext>_</mml:mtext>
<mml:mi>u</mml:mi>
<mml:mi>p</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>k</mml:mi>
<mml:mi>e</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mo>&#x2212;</mml:mo>
<mml:mi>G</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>R</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf158">
<mml:math id="m161">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>o</mml:mi>
<mml:mn>2</mml:mn>
<mml:mtext>_</mml:mtext>
<mml:mi>u</mml:mi>
<mml:mi>p</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>k</mml:mi>
<mml:mi>e</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mo>&#x2212;</mml:mo>
<mml:mi>O</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>R</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf159">
<mml:math id="m162">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>a</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>p</mml:mi>
<mml:mtext>_</mml:mtext>
<mml:mi>m</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mi>N</mml:mi>
<mml:mi>G</mml:mi>
<mml:mi>A</mml:mi>
<mml:mi>M</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf160">
<mml:math id="m163">
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>P</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi mathvariant="bold-italic">m</mml:mi>
<mml:mi mathvariant="bold-italic">a</mml:mi>
<mml:mi mathvariant="bold-italic">x</mml:mi>
<mml:mspace width="0.3333em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf161">
<mml:math id="m164">
<mml:mspace width="2em"/>
<mml:mi mathvariant="bold-italic">s</mml:mi>
<mml:mo>.</mml:mo>
<mml:mi mathvariant="bold-italic">t</mml:mi>
<mml:mo>.</mml:mo>
<mml:mspace width="0.3333em"/>
<mml:mo movablelimits="false" form="prefix">&#x2211;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>S</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>0</mml:mn>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf162">
<mml:math id="m165">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>L</mml:mi>
<mml:msub>
<mml:mrow>
<mml:mi>B</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2264;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>U</mml:mi>
<mml:msub>
<mml:mrow>
<mml:mi>B</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf163">
<mml:math id="m166">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>c</mml:mi>
<mml:mtext>_</mml:mtext>
<mml:mi>u</mml:mi>
<mml:mi>p</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>k</mml:mi>
<mml:mi>e</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mo>&#x2212;</mml:mo>
<mml:mi>G</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>R</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf164">
<mml:math id="m167">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>o</mml:mi>
<mml:mn>2</mml:mn>
<mml:mtext>_</mml:mtext>
<mml:mi>u</mml:mi>
<mml:mi>p</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>k</mml:mi>
<mml:mi>e</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mo>&#x2212;</mml:mo>
<mml:mi>O</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>R</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf165">
<mml:math id="m168">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>a</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>p</mml:mi>
<mml:mtext>_</mml:mtext>
<mml:mi>m</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mi>N</mml:mi>
<mml:mi>G</mml:mi>
<mml:mi>A</mml:mi>
<mml:mi>M</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf166">
<mml:math id="m169">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:msubsup>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>m</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msubsup>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf167">
<mml:math id="m170">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>S</mml:mi>
<mml:mi>F</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mo movablelimits="false" form="prefix">&#x2211;</mml:mo>
<mml:mo stretchy="false">&#x7c;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>j</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo stretchy="false">&#x7c;</mml:mo>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf168">
<mml:math id="m171">
<mml:mi>c</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mn>1,1,1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf169">
<mml:math id="m172">
<mml:mi>n</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<bold>while</bold> <inline-formula id="inf170">
<mml:math id="m173">
<mml:mi>t</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>u</mml:mi>
<mml:mi>e</mml:mi>
</mml:math>
</inline-formula>&#x20;<bold>do</bold>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf171">
<mml:math id="m174">
<mml:mspace width="1em"/>
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:msup>
<mml:mrow>
<mml:mn>2</mml:mn>
</mml:mrow>
<mml:mrow>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msup>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf172">
<mml:math id="m175">
<mml:mspace width="1em"/>
<mml:mi>d</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>z</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>s</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>P</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>P</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>P</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>&#x2003;<bold>for</bold> <inline-formula id="inf173">
<mml:math id="m176">
<mml:mi>i</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
<mml:mspace width="0.3333em"/>
<mml:mi>t</mml:mi>
<mml:mi>o</mml:mi>
<mml:mspace width="0.3333em"/>
<mml:mi>P</mml:mi>
<mml:mo>/</mml:mo>
<mml:mn>2</mml:mn>
</mml:math>
</inline-formula>&#x20;<bold>do</bold>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2003;<bold>for</bold> <inline-formula id="inf174">
<mml:math id="m177">
<mml:mi>j</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
<mml:mspace width="0.3333em"/>
<mml:mi>t</mml:mi>
<mml:mi>o</mml:mi>
<mml:mspace width="0.3333em"/>
<mml:mi>P</mml:mi>
<mml:mo>/</mml:mo>
<mml:mn>2</mml:mn>
</mml:math>
</inline-formula>&#x20;<bold>do</bold>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2003;&#x2003;<bold>for</bold> <inline-formula id="inf175">
<mml:math id="m178">
<mml:mi>k</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
<mml:mspace width="0.3333em"/>
<mml:mi>t</mml:mi>
<mml:mi>o</mml:mi>
<mml:mspace width="0.3333em"/>
<mml:mi>P</mml:mi>
<mml:mo>/</mml:mo>
<mml:mn>2</mml:mn>
</mml:math>
</inline-formula>&#x20;<bold>do</bold>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf176">
<mml:math id="m179">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>d</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mn>2</mml:mn>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mn>2</mml:mn>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mn>2</mml:mn>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>c</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf177">
<mml:math id="m180">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>d</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mn>2</mml:mn>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mn>2</mml:mn>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mn>2</mml:mn>
<mml:mi>k</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>c</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf178">
<mml:math id="m181">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>d</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mn>2</mml:mn>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mn>2</mml:mn>
<mml:mi>j</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1,2</mml:mn>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>c</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf179">
<mml:math id="m182">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>d</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mn>2</mml:mn>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mn>2</mml:mn>
<mml:mi>j</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1,2</mml:mn>
<mml:mi>k</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>c</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf180">
<mml:math id="m183">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>d</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mn>2</mml:mn>
<mml:mi>i</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1,2</mml:mn>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mn>2</mml:mn>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>c</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf181">
<mml:math id="m184">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>d</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mn>2</mml:mn>
<mml:mi>i</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1,2</mml:mn>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mn>2</mml:mn>
<mml:mi>k</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>c</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf182">
<mml:math id="m185">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>d</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mn>2</mml:mn>
<mml:mi>i</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1,2</mml:mn>
<mml:mi>j</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1,2</mml:mn>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>c</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf183">
<mml:math id="m186">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>d</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mn>2</mml:mn>
<mml:mi>i</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1,2</mml:mn>
<mml:mi>j</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1,2</mml:mn>
<mml:mi>k</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>c</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2003;&#x2009;&#x2009;&#x2009;<bold>end&#x20;for</bold>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2009;&#x2009;<bold>end&#x20;for</bold>
</p>
</list-item>
<list-item>
<p>
<bold>end&#x20;for</bold>
</p>
</list-item>
<list-item>
<p>
<bold>for</bold> <inline-formula id="inf184">
<mml:math id="m187">
<mml:mi>i</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
<mml:mspace width="0.3333em" class="nbsp"/>
<mml:mi>t</mml:mi>
<mml:mi>o</mml:mi>
<mml:mspace width="0.3333em" class="nbsp"/>
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula>&#x20;<bold>do</bold>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf185">
<mml:math id="m188">
<mml:mspace width="1em"/>
<mml:mi>b</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>m</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>L</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>G</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>P</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf186">
<mml:math id="m189">
<mml:mspace width="1em"/>
<mml:mi>b</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>m</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>G</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>P</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>i</mml:mi>
</mml:math>
</inline-formula>&#x3e;</p>
</list-item>
<list-item>
<p>
<bold>for</bold> <inline-formula id="inf187">
<mml:math id="m190">
<mml:mi>j</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
<mml:mspace width="0.3333em" class="nbsp"/>
<mml:mi>t</mml:mi>
<mml:mi>o</mml:mi>
<mml:mspace width="0.3333em" class="nbsp"/>
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula>&#x20;<bold>do</bold>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf188">
<mml:math id="m191">
<mml:mspace width="2em"/>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>L</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>P</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>P</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>j</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf189">
<mml:math id="m192">
<mml:mspace width="2em"/>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>P</mml:mi>
<mml:mi>R</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>P</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>j</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<bold>for</bold> <inline-formula id="inf190">
<mml:math id="m193">
<mml:mi>k</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
<mml:mspace width="0.3333em" class="nbsp"/>
<mml:mi>t</mml:mi>
<mml:mi>o</mml:mi>
<mml:mspace width="0.3333em" class="nbsp"/>
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula>&#x20;<bold>do</bold>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf191">
<mml:math id="m194">
<mml:mspace width="1em"/>
<mml:mspace width="2em"/>
<mml:mi>s</mml:mi>
<mml:mi>u</mml:mi>
<mml:mi>m</mml:mi>
<mml:mi>f</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>u</mml:mi>
<mml:mi>x</mml:mi>
<mml:mi>L</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>2</mml:mn>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>S</mml:mi>
<mml:mi>F</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>P</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>k</mml:mi>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf192">
<mml:math id="m195">
<mml:mspace width="1em"/>
<mml:mspace width="2em"/>
<mml:mi>s</mml:mi>
<mml:mi>u</mml:mi>
<mml:mi>n</mml:mi>
<mml:mi>f</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>u</mml:mi>
<mml:mi>x</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>2</mml:mn>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>T</mml:mi>
<mml:mi>M</mml:mi>
<mml:mi>S</mml:mi>
<mml:mi>F</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>P</mml:mi>
<mml:mo>&#xd7;</mml:mo>
<mml:mi>k</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<bold>if</bold> <inline-formula id="inf193">
<mml:math id="m196">
<mml:mi>d</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>0</mml:mn>
</mml:math>
</inline-formula> t<bold>hen</bold>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf194">
<mml:math id="m197">
<mml:mi mathvariant="bold-italic">c</mml:mi>
<mml:mi mathvariant="bold-italic">o</mml:mi>
<mml:mi mathvariant="bold-italic">n</mml:mi>
<mml:mi mathvariant="bold-italic">t</mml:mi>
<mml:mi mathvariant="bold-italic">i</mml:mi>
<mml:mi mathvariant="bold-italic">n</mml:mi>
<mml:mi mathvariant="bold-italic">u</mml:mi>
<mml:mi mathvariant="bold-italic">e</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<bold>else</bold>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf195">
<mml:math id="m198">
<mml:mi mathvariant="bold-italic">m</mml:mi>
<mml:mi mathvariant="bold-italic">i</mml:mi>
<mml:mi mathvariant="bold-italic">n</mml:mi>
<mml:mspace width="0.3333em"/>
<mml:mo movablelimits="false" form="prefix">&#x2211;</mml:mo>
<mml:mo stretchy="false">&#x7c;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo stretchy="false">&#x7c;</mml:mo>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf196">
<mml:math id="m199">
<mml:mspace width="2em"/>
<mml:mi mathvariant="bold-italic">s</mml:mi>
<mml:mo>.</mml:mo>
<mml:mi mathvariant="bold-italic">t</mml:mi>
<mml:mo>.</mml:mo>
<mml:mspace width="0.3333em"/>
<mml:mo movablelimits="false" form="prefix">&#x2211;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>S</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>a</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>0</mml:mn>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf197">
<mml:math id="m200">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>L</mml:mi>
<mml:msub>
<mml:mrow>
<mml:mi>B</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2264;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>U</mml:mi>
<mml:msub>
<mml:mrow>
<mml:mi>B</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf198">
<mml:math id="m201">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>c</mml:mi>
<mml:mtext>_</mml:mtext>
<mml:mi>u</mml:mi>
<mml:mi>p</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>k</mml:mi>
<mml:mi>e</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mo>&#x2212;</mml:mo>
<mml:mi>G</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>R</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf199">
<mml:math id="m202">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>o</mml:mi>
<mml:mn>2</mml:mn>
<mml:mtext>_</mml:mtext>
<mml:mi>u</mml:mi>
<mml:mi>p</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>k</mml:mi>
<mml:mi>e</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mo>&#x2212;</mml:mo>
<mml:mi>O</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>R</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf200">
<mml:math id="m203">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>a</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>p</mml:mi>
<mml:mtext>_</mml:mtext>
<mml:mi>m</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mi>N</mml:mi>
<mml:mi>G</mml:mi>
<mml:mi>A</mml:mi>
<mml:mi>M</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf201">
<mml:math id="m204">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>b</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>m</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>L</mml:mi>
<mml:mi>U</mml:mi>
<mml:mo>&#x2264;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>b</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>m</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>B</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf202">
<mml:math id="m205">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>L</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>&#x2264;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>B</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf203">
<mml:math id="m206">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>s</mml:mi>
<mml:mi>u</mml:mi>
<mml:mi>m</mml:mi>
<mml:mi>f</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>u</mml:mi>
<mml:mi>x</mml:mi>
<mml:mi>L</mml:mi>
<mml:mi>B</mml:mi>
<mml:mo>&#x2264;</mml:mo>
<mml:mo movablelimits="false" form="prefix">&#x2211;</mml:mo>
<mml:mo stretchy="false">&#x7c;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo stretchy="false">&#x7c;</mml:mo>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>s</mml:mi>
<mml:mi>u</mml:mi>
<mml:mi>m</mml:mi>
<mml:mi>f</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>u</mml:mi>
<mml:mi>x</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>B</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf204">
<mml:math id="m207">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>K</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">{</mml:mo>
<mml:mrow>
<mml:msub>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo stretchy="false">&#x7c;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x3c;</mml:mo>
<mml:mn>1</mml:mn>
<mml:msup>
<mml:mrow>
<mml:mn>0</mml:mn>
</mml:mrow>
<mml:mrow>
<mml:mo>&#x2212;</mml:mo>
<mml:mn>5</mml:mn>
</mml:mrow>
</mml:msup>
</mml:mrow>
<mml:mo stretchy="false">}</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<bold>If</bold> <inline-formula id="inf205">
<mml:math id="m208">
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>e</mml:mi>
<mml:mspace width="0.3333em"/>
<mml:mi>f</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>t</mml:mi>
<mml:mspace width="0.3333em"/>
<mml:mi>L</mml:mi>
<mml:mi>P</mml:mi>
<mml:mspace width="0.3333em"/>
<mml:mi>i</mml:mi>
<mml:mi>s</mml:mi>
<mml:mspace width="0.3333em"/>
<mml:mi>i</mml:mi>
<mml:mi>n</mml:mi>
<mml:mi>f</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>b</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>e</mml:mi>
</mml:math>
</inline-formula>&#x20;<bold>then</bold>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2003;&#x2003;&#x2003;&#x2003;&#x2009;&#x2009;<inline-formula id="inf206">
<mml:math id="m209">
<mml:mi>K</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>&#x3d5;</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2003;&#x2003;&#x2003;&#x2003;&#x2009;&#x2009;<inline-formula id="inf207">
<mml:math id="m210">
<mml:mi>P</mml:mi>
<mml:mi>R</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>0</mml:mn>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2003;&#x2003;&#x2003;&#x2003;&#x2009;&#x2009;<inline-formula id="inf208">
<mml:math id="m211">
<mml:mi>c</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>0</mml:mn>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<bold>else</bold>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf209">
<mml:math id="m212">
<mml:mi mathvariant="bold-italic">m</mml:mi>
<mml:mi mathvariant="bold-italic">a</mml:mi>
<mml:mi mathvariant="bold-italic">x</mml:mi>
<mml:mspace width="0.3333em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf210">
<mml:math id="m213">
<mml:mspace width="2em"/>
<mml:mi mathvariant="bold-italic">s</mml:mi>
<mml:mo>.</mml:mo>
<mml:mi mathvariant="bold-italic">t</mml:mi>
<mml:mo>.</mml:mo>
<mml:mspace width="0.3333em"/>
<mml:mo movablelimits="false" form="prefix">&#x2211;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>S</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>a</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>0</mml:mn>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf211">
<mml:math id="m214">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:mi>L</mml:mi>
<mml:msub>
<mml:mrow>
<mml:mi>B</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2264;</mml:mo>
<mml:msub>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2264;</mml:mo>
<mml:mi>U</mml:mi>
<mml:msub>
<mml:mrow>
<mml:mi>B</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula> if <inline-formula id="inf212">
<mml:math id="m215">
<mml:msub>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2209;</mml:mo>
<mml:mi>K</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf213">
<mml:math id="m216">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>0</mml:mn>
<mml:mspace width="0.3333em" class="nbsp"/>
<mml:mi mathvariant="normal">i</mml:mi>
<mml:mi mathvariant="normal">f</mml:mi>
<mml:mspace width="0.3333em" class="nbsp"/>
<mml:msub>
<mml:mrow>
<mml:mi>v</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>b</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2209;</mml:mo>
<mml:mi>K</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf214">
<mml:math id="m217">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>c</mml:mi>
<mml:mtext>_</mml:mtext>
<mml:mi>u</mml:mi>
<mml:mi>p</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>k</mml:mi>
<mml:mi>e</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mo>&#x2212;</mml:mo>
<mml:mi>G</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>R</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf215">
<mml:math id="m218">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>o</mml:mi>
<mml:mn>2</mml:mn>
<mml:mtext>_</mml:mtext>
<mml:mi>u</mml:mi>
<mml:mi>p</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>k</mml:mi>
<mml:mi>e</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mo>&#x2212;</mml:mo>
<mml:mi>O</mml:mi>
<mml:mi>U</mml:mi>
<mml:mi>R</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf216">
<mml:math id="m219">
<mml:mspace width="2em"/>
<mml:mspace width="2em"/>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>a</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>p</mml:mi>
<mml:mtext>_</mml:mtext>
<mml:mi>m</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:mi>N</mml:mi>
<mml:mi>G</mml:mi>
<mml:mi>A</mml:mi>
<mml:mi>M</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf217">
<mml:math id="m220">
<mml:mi>c</mml:mi>
<mml:mrow>
<mml:mo stretchy="false">(</mml:mo>
<mml:mrow>
<mml:mi>i</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>j</mml:mi>
<mml:mo>,</mml:mo>
<mml:mi>k</mml:mi>
</mml:mrow>
<mml:mo stretchy="false">)</mml:mo>
</mml:mrow>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<bold>if</bold> <inline-formula id="inf218">
<mml:math id="m221">
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:msubsup>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>s</mml:mi>
<mml:mi>h</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>l</mml:mi>
<mml:mi>d</mml:mi>
</mml:mrow>
</mml:msubsup>
</mml:math>
</inline-formula> and <inline-formula id="inf219">
<mml:math id="m222">
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
<mml:mo>&#x2265;</mml:mo>
<mml:msubsup>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>m</mml:mi>
<mml:mi>i</mml:mi>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msubsup>
</mml:math>
</inline-formula>&#x20;<bold>then</bold>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2003;&#x2003;&#x2003;&#x2003;&#x2003;&#x2009;&#x2009;<bold>return</bold> <inline-formula id="inf220">
<mml:math id="m223">
<mml:mi>K</mml:mi>
</mml:math>
</inline-formula>, <inline-formula id="inf221">
<mml:math id="m224">
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>t</mml:mi>
<mml:mi>a</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>g</mml:mi>
<mml:mi>e</mml:mi>
<mml:mi>t</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula>, <inline-formula id="inf222">
<mml:math id="m225">
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>g</mml:mi>
<mml:mi>r</mml:mi>
<mml:mi>o</mml:mi>
<mml:mi>w</mml:mi>
<mml:mi>t</mml:mi>
<mml:mi>h</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2003;&#x2003;&#x2003;&#x2003;&#x2009;&#x2009;&#x2009;<bold>end&#x20;if</bold>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2003;&#x2003;&#x2003;&#x2009;&#x2009;&#x2009;<bold>end&#x20;if</bold>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2003;&#x2003;&#x2009;&#x2009;<bold>end&#x20;if</bold>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2003;&#x2009;&#x2009;<bold>end&#x20;for</bold>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2009;<bold>end&#x20;for</bold>
</p>
</list-item>
<list-item>
<p>&#x2009;<bold>end&#x20;for</bold>
</p>
</list-item>
<list-item>
<p>
<bold>if</bold> <inline-formula id="inf223">
<mml:math id="m226">
<mml:mi>n</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>m</mml:mi>
<mml:mi>a</mml:mi>
<mml:msub>
<mml:mrow>
<mml:mi>x</mml:mi>
</mml:mrow>
<mml:mrow>
<mml:mi>n</mml:mi>
</mml:mrow>
</mml:msub>
</mml:math>
</inline-formula> t<bold>hen</bold>
</p>
</list-item>
<list-item>
<p>&#x2003;&#x2009;&#x2009;<inline-formula id="inf224">
<mml:math id="m227">
<mml:mi mathvariant="bold-italic">b</mml:mi>
<mml:mi mathvariant="bold-italic">r</mml:mi>
<mml:mi mathvariant="bold-italic">e</mml:mi>
<mml:mi mathvariant="bold-italic">a</mml:mi>
<mml:mi mathvariant="bold-italic">k</mml:mi>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<bold>end&#x20;if</bold>
</p>
</list-item>
<list-item>
<p>
<inline-formula id="inf225">
<mml:math id="m228">
<mml:mi>n</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:mi>n</mml:mi>
<mml:mo>&#x2b;</mml:mo>
<mml:mn>1</mml:mn>
</mml:math>
</inline-formula>
</p>
</list-item>
<list-item>
<p>
<bold>end&#x20;while</bold>
</p>
</list-item>
</list>
</p>
</statement>
</p>
</sec>
<sec id="s3-3">
<title>3.3 Computational Experiments</title>
<p>The dataset used in the computational experiments was iJO1366, which is a GSMM of <italic>Escherichia coli</italic> K-12 MG1655 from the BiGG database with 1805 metabolites and 2583 reactions (<xref ref-type="bibr" rid="B11">Orth et&#x20;al., 2011</xref>; <xref ref-type="bibr" rid="B5">King et&#x20;al., 2016</xref>). All procedures of DynCubeProd were implemented based on Gurobi, COBRA Toolbox and MATLAB on a Windows machine with Intel(R) Core(TM) i5-8500 CPU 3.00&#xa0;GHz 6-core processor and 32.0&#xa0;GB&#x20;RAM.</p>
<p>If the target metabolite is not connected to an external reaction, then, an auxiliary external reaction is added, and the growth coupling is evaluated by GR and the outgoing flux from the additional external reaction, which is also called&#x20;PR.</p>
<p>
<xref ref-type="fig" rid="F3">Figure&#x20;3A</xref> shows the computing time of DynCubeProd and CubeProd when applied to iJO1366 under aerobic conditions at different values of P. It also shows the ratio of the number of success cases to the number of target metabolites. For <inline-formula id="inf226">
<mml:math id="m229">
<mml:mi>P</mml:mi>
<mml:mo>&#x2265;</mml:mo>
<mml:mn>16</mml:mn>
</mml:math>
</inline-formula>, the reaction deletion strategies were obtained for more than 95% of the target metabolites. It should be noted that the success ratio of DynCubeProd and CubeProd is always the same for the same P. The computing time for DynCubeProd with <italic>p</italic>&#x20;&#x3d; 32 was only approximately quarter of the computing time of CubeProd with <italic>p</italic>&#x20;&#x3d; 16. <xref ref-type="fig" rid="F3">Figure&#x20;3B</xref> visually compares the computing time increase by <inline-formula id="inf227">
<mml:math id="m230">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> between DynCubeProd and CubeProd.</p>
<fig id="F3" position="float">
<label>FIGURE 3</label>
<caption>
<p>
<bold>(A)</bold> Computation time and success ratio when DynCubeProd and CubeProd were applied to iJO1366 under aerobic conditions for different values of <inline-formula id="inf228">
<mml:math id="m231">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula>. <bold>(B)</bold> Visual comparison of the computation time of DynCubeProd and CubeProd of (A).</p>
</caption>
<graphic xlink:href="fbinf-01-716112-g003.tif"/>
</fig>
<p>Furthermore, under anaerobic conditions of iJO1366, DynCubeProd succeeded in computing the reaction deletion strategies for 76 of the 211 target metabolites for which the EFV-based method (<xref ref-type="bibr" rid="B17">von Kamp and Klamt, 2017</xref>), which is one of the best methods, could&#x20;not.</p>
</sec>
</sec>
<sec id="s4">
<title>4 Discussion</title>
<p>To date, numerous algorithms have been proposed using constraint-based models to compute reaction deletion strategies for growth coupling in GSMM. In particular, the EFV-based method is one of the most efficient methods to determine reaction deletion strategies that achieve growth coupling of target metabolites. The success ratios of the EFV-based method for iJO1366 under aerobic and anaerobic conditions were 99.4 and 77.4%, respectively (<xref ref-type="bibr" rid="B17">von Kamp and Klamt, 2017</xref>). Because the success ratio of DynCubeProd for iJO1366 under aerobic conditions was 95.99% for <italic>p</italic>&#x20;&#x3d; 32, the EFV-based method is more efficient under aerobic conditions and there is almost no room for improvement.</p>
<p>However, for iJO1366 under anaerobic conditions, the success ratio of the EFV-based method is 77.4%. An effective method for target metabolites for which reaction deletion strategies could not be calculated by the EFV-based method must be developed. Because DynCubeProd succeeded in determining the reaction deletion strategies for 76 of such 211 target metabolites as described in <xref ref-type="sec" rid="s3">Section 3</xref>, we can conclude that DynCubeProd can play a complementary role to the EFV-based method in the calculation of reaction deletion strategies for growth coupling.</p>
<p>The success ratio of DynCubeProd and CubeProd is always the same if <inline-formula id="inf229">
<mml:math id="m232">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> is the same. However, since DynCubeProd is faster than CubeProd, DynCubeProd can reach a larger <inline-formula id="inf230">
<mml:math id="m233">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> in realistic computation time. Therefore, if the allowable computation time is the same, DynCubeProd is more capable of computing reaction deletion strategies than CubeProd. The speed-up of the computation is achieved by not examining the constraint space where no solution exists, and by stopping the algorithm as soon as a solution satisfying the requirements is obtained. Additionally, CubeProd is a method developed by (<xref ref-type="bibr" rid="B15">Tamura, 2021a</xref>); however in this study, the definition of TMSF is slightly modified to be the absolute sum of reaction rates at PR maximization.</p>
<p>If the computation is conducted for very large <inline-formula id="inf231">
<mml:math id="m234">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:msup>
<mml:mrow>
<mml:mn>2</mml:mn>
</mml:mrow>
<mml:mrow>
<mml:mi>i</mml:mi>
</mml:mrow>
</mml:msup>
</mml:math>
</inline-formula>, the computation time of DynCubeProd becomes too long. Therefore, values up to <inline-formula id="inf232">
<mml:math id="m235">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
<mml:msup>
<mml:mrow>
<mml:mn>2</mml:mn>
</mml:mrow>
<mml:mrow>
<mml:mi>i</mml:mi>
</mml:mrow>
</mml:msup>
<mml:mo>&#x3d;</mml:mo>
<mml:mn>32</mml:mn>
</mml:math>
</inline-formula> were used in the computational experiments in this study. Even if the computation is conducted for all <inline-formula id="inf233">
<mml:math id="m236">
<mml:mi>P</mml:mi>
<mml:mo>&#x3d;</mml:mo>
</mml:math>
</inline-formula> 2, 4, 8, 16, and 32, the total number of cubes can be regarded as not too large to conduct several LPs for each cube and as an ignorable constant when estimating the computational complexity of DynCubeProd. It is known that an LP can be solved in a time proportional to the polynomial function of the number of variables (<xref ref-type="bibr" rid="B1">Arora and Barak, 2009</xref>). Because DynCubeProd solves several LPs for the entire problem and for each cube, the computational complexity of the algorithm is proportional to the polynomial function of the problem size because <inline-formula id="inf234">
<mml:math id="m237">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> can be regarded as a constant.</p>
<sec id="s4-1">
<title>4.1 Comparison With GridProd and CubeProd</title>
<p>DynCubeProd was developed as an extension of CubeProd that was developed as an extension of GridProd. Therefore, a problem example that can be solved with CubeProd, but not with GridProd, may help understand for DynCubeProd. Such an example is illustrated&#x20;below.</p>
<p>GridProd picks two-dimensional limits and grids the constraint space to solve the problem by using bilevel optimization approach (<xref ref-type="bibr" rid="B14">Tamura, 2018</xref>). <xref ref-type="table" rid="T2">Table&#x20;2A</xref> shows example results of GridProd applied on the toy network. GR and PR are selected as the two-dimensional limits of GirdPord, the gridding of the GridProd will be performed on the axes of these two constraints, and the size of the gridding depends on the designated value of <inline-formula id="inf235">
<mml:math id="m238">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula>. In <xref ref-type="table" rid="T2">Table&#x20;2A</xref>, the second and third rows show results of the bilevel optimization approach of a grid with <italic>p</italic>&#x20;&#x3d; 4, that is, GR <inline-formula id="inf236">
<mml:math id="m239">
<mml:mo>&#x2208;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1.5</mml:mn>
<mml:mo>,</mml:mo>
<mml:mspace width="0.3333em"/>
<mml:mn>2.25</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula> and PR <inline-formula id="inf237">
<mml:math id="m240">
<mml:mo>&#x2208;</mml:mo>
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0.75</mml:mn>
<mml:mo>,</mml:mo>
<mml:mspace width="0.3333em"/>
<mml:mn>1.5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>. The grid becomes a point when <inline-formula id="inf238">
<mml:math id="m241">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> is infinite. The fourth and fifth rows represent the results in the point of optimal solution when <inline-formula id="inf239">
<mml:math id="m242">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> is infinite. The fifth row shows that the acceptable solution cannot be obtained even when <inline-formula id="inf240">
<mml:math id="m243">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> is under infinity.</p>
<table-wrap id="T2" position="float">
<label>TABLE 2</label>
<caption>
<p>
<bold>(A)</bold> GridProd cannot find reaction deletion strategies that achieve growth coupling for the toy example of <xref ref-type="fig" rid="F2">Figure&#x20;2A</xref> even when <inline-formula id="inf241">
<mml:math id="m244">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> is infinity. <bold>(B)</bold> However, CubeProd can find reaction deletion strategies for the same problem</p>
</caption>
<table>
<thead valign="top">
<tr>
<th colspan="10" align="left">(A)</th>
</tr>
<tr>
<th align="left">PR</th>
<th align="center">GR</th>
<th align="left"/>
<th align="center">R1</th>
<th align="center">R2</th>
<th align="center">R3</th>
<th align="center">R4</th>
<th align="center">R5</th>
<th align="center">R6</th>
<th align="center">R7</th>
</tr>
</thead>
<tbody valign="top">
<tr>
<td rowspan="2" align="left">
<inline-formula id="inf242">
<mml:math id="m245">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1.5</mml:mn>
<mml:mo>,</mml:mo>
<mml:mn>2.25</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf243">
<mml:math id="m246">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0.75</mml:mn>
<mml:mo>,</mml:mo>
<mml:mn>1.5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="left">1st LP</td>
<td align="center">3</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
<td align="center">0</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
</tr>
<tr>
<td align="left">2nd LP</td>
<td align="center">3</td>
<td align="center">0</td>
<td align="center">3</td>
<td align="center">0</td>
<td align="center">0</td>
<td align="center">3</td>
<td align="center">0</td>
</tr>
<tr>
<td rowspan="2" align="left">
<inline-formula id="inf244">
<mml:math id="m247">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>2,2</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf245">
<mml:math id="m248">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1,1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="left">1st LP</td>
<td align="center">3</td>
<td align="center">2</td>
<td align="center">1</td>
<td align="center">2</td>
<td align="center">0</td>
<td align="center">1</td>
<td align="center">2</td>
</tr>
<tr>
<td align="left">2nd LP</td>
<td align="center">3</td>
<td align="center">0</td>
<td align="center">3</td>
<td align="center">0</td>
<td align="center">0</td>
<td align="center">3</td>
<td align="center">0</td>
</tr>
<tr>
<td colspan="10" align="left">
<bold>(B)</bold>
</td>
</tr>
<tr>
<td align="left">
<bold>PR</bold>
</td>
<td align="center">
<bold>GR</bold>
</td>
<td align="center">
<bold>SF</bold>
</td>
<td align="center">
<bold>R1</bold>
</td>
<td align="center">
<bold>R2</bold>
</td>
<td align="center">
<bold>R3</bold>
</td>
<td align="center">
<bold>R4</bold>
</td>
<td align="center">
<bold>R5</bold>
</td>
<td align="center">
<bold>R6</bold>
</td>
<td align="center">
<bold>R7</bold>
</td>
</tr>
<tr>
<td rowspan="2" align="left">
<inline-formula id="inf246">
<mml:math id="m249">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1.5</mml:mn>
<mml:mo>,</mml:mo>
<mml:mn>2.25</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf247">
<mml:math id="m250">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>0.75</mml:mn>
<mml:mo>,</mml:mo>
<mml:mn>1.5</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf248">
<mml:math id="m251">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>12,14</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">3</td>
<td align="center">2.25</td>
<td align="center">0.75</td>
<td align="center">2.25</td>
<td align="center">0.75</td>
<td align="center">1.5</td>
<td align="center">1.5</td>
</tr>
<tr>
<td align="center">3</td>
<td align="center">0</td>
<td align="center">3</td>
<td align="center">0</td>
<td align="center">0</td>
<td align="center">3</td>
<td align="center">0</td>
</tr>
<tr>
<td rowspan="2" align="left">
<inline-formula id="inf249">
<mml:math id="m252">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>2,2</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf250">
<mml:math id="m253">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>1,1</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td rowspan="2" align="center">
<inline-formula id="inf251">
<mml:math id="m254">
<mml:mrow>
<mml:mo stretchy="false">[</mml:mo>
<mml:mrow>
<mml:mn>13,13</mml:mn>
</mml:mrow>
<mml:mo stretchy="false">]</mml:mo>
</mml:mrow>
</mml:math>
</inline-formula>
</td>
<td align="center">3</td>
<td align="center">3</td>
<td align="center">0</td>
<td align="center">3</td>
<td align="center">1</td>
<td align="center">1</td>
<td align="center">2</td>
</tr>
<tr>
<td align="center">3</td>
<td align="center">3</td>
<td align="center">0</td>
<td align="center">3</td>
<td align="center">1</td>
<td align="center">1</td>
<td align="center">2</td>
</tr>
</tbody>
</table>
</table-wrap>
<p>CubeProd, which is an extension of GridProd, divides the entire constraint space into small cubes (<xref ref-type="bibr" rid="B15">Tamura, 2021a</xref>). <xref ref-type="table" rid="T2">Table&#x20;2B</xref> depicts a sample result of applying CubeProd to the toy network. The second to fifth rows show the results of bilevel optimization for cubes containing the optimal solution for different values of P. Since GR &#x3d; 3 and PR &#x3d; 0 are obtained, CubeProd cannot obtain the reaction deletion strategy for growth coupling for PR<inline-formula id="inf253">
<mml:math id="m256">
<mml:mo>&#x2208;</mml:mo>
</mml:math>
</inline-formula>[1.5,2.25], GR<inline-formula id="inf254">
<mml:math id="m257">
<mml:mo>&#x2208;</mml:mo>
</mml:math>
</inline-formula>[0.75,1.5], and SF<inline-formula id="inf255">
<mml:math id="m258">
<mml:mo>&#x2208;</mml:mo>
</mml:math>
</inline-formula>[12,14]. However, since GR &#x3d; 1 and PR &#x3d; 2 are obtained, CubeProd can obtain the reaction deletion strategy for growth coupling for PR<inline-formula id="inf256">
<mml:math id="m259">
<mml:mo>&#x2208;</mml:mo>
</mml:math>
</inline-formula>[2,2], GR<inline-formula id="inf257">
<mml:math id="m260">
<mml:mo>&#x2208;</mml:mo>
</mml:math>
</inline-formula>[1,1], and SF<inline-formula id="inf258">
<mml:math id="m261">
<mml:mo>&#x2208;</mml:mo>
</mml:math>
</inline-formula>[13,13]. As shown in this example, the larger the value of <inline-formula id="inf259">
<mml:math id="m262">
<mml:mi>P</mml:mi>
</mml:math>
</inline-formula> and smaller the cube sizes, the easier it is to find a solution.</p>
</sec>
<sec id="s4-2">
<title>4.2 Other Methods</title>
<p>For the problem of calculating reaction-deletion strategies for growth coupling of the constraint-based models, there are various methods other than those introduced so far. OptGene is a computational method that uses bio-inspired algorithms to optimize gene deletion sets (<xref ref-type="bibr" rid="B12">Patil et&#x20;al., 2005</xref>). Genetic Design through Local Search (GDLS) was developed to&#x20;use global optimal search to find genetic design, and compared with heuristic search based on evolutionary algorithm and simulated annealing, GDLS performs well (<xref ref-type="bibr" rid="B7">Lun et&#x20;al., 2009</xref>). EMILiO uses iterative linear programs (<xref ref-type="bibr" rid="B18">Yang et&#x20;al., 2011</xref>). FastPros evaluates the potential of specific reaction knockout to produce specific metabolites by shadow pricing the constraints in FBA, thereby&#x20;generating a screening score to obtain candidate knockout sets (<xref ref-type="bibr" rid="B10">Ohno et&#x20;al., 2014</xref>). IdealKnock can effectively evaluate the production potential of different biochemical products in the system, just by knocking out some pathways and combining with the OptKnock or OptGene framework (<xref ref-type="bibr" rid="B4">Gu et&#x20;al., 2016</xref>). Parsimonious enzyme usage FBA (pFBA) not only used the metabolic network, but also used proteomics and transcriptomics data to confirm that almost all path dosages predicted by the FBA optimization method were consistent (<xref ref-type="bibr" rid="B6">Lewis et&#x20;al., 2010</xref>). IdealKnock utilizes the concept of ideal-type flux distribution (<xref ref-type="bibr" rid="B4">Gu et&#x20;al., 2016</xref>). PSOMCS also uses the perspective of EFM, combined with particle swarm optimization algorithms, to obtain an optimal design that meets multiple goals (<xref ref-type="bibr" rid="B9">Nair et&#x20;al., 2017</xref>).</p>
<p>For the performance evaluation of DynCubeProd, it is reasonable to compare it with GridProd and CubeProd for the evaluation of computation time reduction and algorithm behavior comparison, and it is reasonable to compare it with the EFV-based method, a method with the best success rate among existing methods, for the evaluation of success rate. Furthermore, neither OptKnock nor GDLS, one of the most standard and popular methods, could determine reaction deletion strategies for any target metabolite within 1&#xa0;hour. Therefore, although there are many other existing methods as mentioned above, it is reasonable to narrow down the comparison targets of DynCubeProd to GridProd, CubeProd, and EFV methods.</p>
</sec>
<sec id="s5">
<title>4.3 Conclusion</title>
<p>DynCubeProd is an improved version of CubeProd, which is an existing algorithm based on solution space decomposition. The improvements in DynCubeProd are as follows. 1) While CubeProd divides the solution space based on pre-specified parameters, DynCubeProd gradually divides the solution space into smaller and smaller pieces. 2) CubeProd mechanically explores even the subspace where no solution is expected to exist, while DynCubeProd stops dividing the solution space when no solution exists. 3) While CubeProd searches the entire solution space, DynCubeProd stops as soon&#x20;as it finds a solution that meets the conditions. The&#x20;results of computer experiments using iJO1366 confirmed that DynCubeProd reduces the computation time more than 10&#x20;times than CubeProd. The reduction in computation time enabled finer solution space partitioning, and reaction deletion strategies could be calculated for about&#x20;40% of the target metabolites for which reaction deletion strategies could not be obtained by the EFV-based method. In this study, we developed DynCubeProd, by improving the computation speed of CubeProd, which enabled us to calculate reaction deletion strategies in anaerobic conditions for many target compounds that could not be calculated before.</p>
</sec>
</sec>
</body>
<back>
<sec id="s6">
<title>Data Availability Statement</title>
<p>The developed software is available on <ext-link ext-link-type="uri" xlink:href="https://github.com/Ma-Yier/DynCubeProd">https://github.com/Ma-Yier/DynCubeProd</ext-link>.</p>
</sec>
<sec id="s7">
<title>Author Contributions</title>
<p>YM implemented the source codes and conducted computational experiments. TT designed the study. The manuscript was written both by YM and&#x20;TT.</p>
</sec>
<sec id="s8">
<title>Funding</title>
<p>This study was partially supported by grants from JSPS, KAKENHI &#x23;20H04242.</p>
</sec>
<sec sec-type="COI-statement" id="s9">
<title>Conflict of Interest</title>
<p>The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.</p>
</sec>
<sec sec-type="disclaimer" id="s10">
<title>Publisher&#x2019;s Note</title>
<p>All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.</p>
</sec>
<ack>
<p>We appreciate all members of the mathematical bioinformatics group of Bioinformatics Center, Institute for Chemical Research, Kyoto University.</p>
</ack>
<fn-group>
<fn id="fn1">
<label>1</label>
<p>Because it is difficult to determine TMSF in polynomial time, we approximate TMSF by SF when PR is maximized, and [0,2&#x22C5;TMSF] was used instead of [0,TMSF] in the computational experiments and examples.</p>
</fn>
</fn-group>
<ref-list>
<title>References</title>
<ref id="B1">
<citation citation-type="book">
<person-group person-group-type="author">
<name>
<surname>Arora</surname>
<given-names>S.</given-names>
</name>
<name>
<surname>Barak</surname>
<given-names>B.</given-names>
</name>
</person-group> (<year>2009</year>). <source>Computational Complexity: A Modern Approach</source>. <publisher-name>Cambridge University Press</publisher-name>.</citation>
</ref>
<ref id="B2">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Bailey</surname>
<given-names>J.&#x20;E.</given-names>
</name>
</person-group> (<year>1991</year>). <article-title>Toward a Science of Metabolic Engineering</article-title>. <source>Science</source> <volume>252</volume>, <fpage>1668</fpage>&#x2013;<lpage>1675</lpage>. <pub-id pub-id-type="doi">10.1126/science.2047876</pub-id> </citation>
</ref>
<ref id="B3">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Burgard</surname>
<given-names>A. P.</given-names>
</name>
<name>
<surname>Pharkya</surname>
<given-names>P.</given-names>
</name>
<name>
<surname>Maranas</surname>
<given-names>C. D.</given-names>
</name>
</person-group> (<year>2003</year>). <article-title>Optknock: a Bilevel Programming Framework for Identifying Gene Knockout Strategies for Microbial Strain Optimization</article-title>. <source>Biotechnol. Bioeng.</source> <volume>84</volume>, <fpage>647</fpage>&#x2013;<lpage>657</lpage>. <pub-id pub-id-type="doi">10.1002/bit.10803</pub-id> </citation>
</ref>
<ref id="B4">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Gu</surname>
<given-names>D.</given-names>
</name>
<name>
<surname>Zhang</surname>
<given-names>C.</given-names>
</name>
<name>
<surname>Zhou</surname>
<given-names>S.</given-names>
</name>
<name>
<surname>Wei</surname>
<given-names>L.</given-names>
</name>
<name>
<surname>Hua</surname>
<given-names>Q.</given-names>
</name>
</person-group> (<year>2016</year>). <article-title>Idealknock: a Framework for Efficiently Identifying Knockout Strategies Leading to Targeted Overproduction</article-title>. <source>Comput. Biol. Chem.</source> <volume>61</volume>, <fpage>229</fpage>&#x2013;<lpage>237</lpage>. <pub-id pub-id-type="doi">10.1016/j.compbiolchem.2016.02.014</pub-id> </citation>
</ref>
<ref id="B5">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>King</surname>
<given-names>Z. A.</given-names>
</name>
<name>
<surname>Lu</surname>
<given-names>J.</given-names>
</name>
<name>
<surname>Dr&#xe4;ger</surname>
<given-names>A.</given-names>
</name>
<name>
<surname>Miller</surname>
<given-names>P.</given-names>
</name>
<name>
<surname>Federowicz</surname>
<given-names>S.</given-names>
</name>
<name>
<surname>Lerman</surname>
<given-names>J.&#x20;A.</given-names>
</name>
<etal/>
</person-group> (<year>2016</year>). <article-title>Bigg Models: A Platform for Integrating, Standardizing and Sharing Genome-Scale Models</article-title>. <source>Nucleic Acids Res.</source> <volume>44</volume>, <fpage>D515</fpage>&#x2013;<lpage>D522</lpage>. <pub-id pub-id-type="doi">10.1093/nar/gkv1049</pub-id> </citation>
</ref>
<ref id="B6">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Lewis</surname>
<given-names>N. E.</given-names>
</name>
<name>
<surname>Hixson</surname>
<given-names>K. K.</given-names>
</name>
<name>
<surname>Conrad</surname>
<given-names>T. M.</given-names>
</name>
<name>
<surname>Lerman</surname>
<given-names>J.&#x20;A.</given-names>
</name>
<name>
<surname>Charusanti</surname>
<given-names>P.</given-names>
</name>
<name>
<surname>Polpitiya</surname>
<given-names>A. D.</given-names>
</name>
<etal/>
</person-group> (<year>2010</year>). <article-title>Omic data from evolved e. coli are consistent with computed optimal growth from genome-scale models</article-title>. <source>Mol. Syst. Biol.</source> <volume>6</volume>, <fpage>390</fpage>. <pub-id pub-id-type="doi">10.1038/msb.2010.47</pub-id> </citation>
</ref>
<ref id="B7">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Lun</surname>
<given-names>D. S.</given-names>
</name>
<name>
<surname>Rockwell</surname>
<given-names>G.</given-names>
</name>
<name>
<surname>Guido</surname>
<given-names>N. J.</given-names>
</name>
<name>
<surname>Baym</surname>
<given-names>M.</given-names>
</name>
<name>
<surname>Kelner</surname>
<given-names>J.&#x20;A.</given-names>
</name>
<name>
<surname>Berger</surname>
<given-names>B.</given-names>
</name>
<etal/>
</person-group> (<year>2009</year>). <article-title>Large-scale Identification of Genetic Design Strategies Using Local Search</article-title>. <source>Mol. Syst. Biol.</source> <volume>5</volume>, <fpage>296</fpage>. <pub-id pub-id-type="doi">10.1038/msb.2009.57</pub-id> </citation>
</ref>
<ref id="B8">
<citation citation-type="book">
<person-group person-group-type="author">
<name>
<surname>Maranas</surname>
<given-names>C. D.</given-names>
</name>
<name>
<surname>Zomorrodi</surname>
<given-names>A. R.</given-names>
</name>
</person-group> (<year>2016</year>). <source>Optimization Methods in Metabolic Networks</source>. <publisher-name>John Wiley &#x26; Sons</publisher-name>.</citation>
</ref>
<ref id="B9">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Nair</surname>
<given-names>G.</given-names>
</name>
<name>
<surname>Jungreuthmayer</surname>
<given-names>C.</given-names>
</name>
<name>
<surname>Zanghellini</surname>
<given-names>J.</given-names>
</name>
</person-group> (<year>2017</year>). <article-title>Optimal Knockout Strategies in Genome-Scale Metabolic Networks Using Particle Swarm Optimization</article-title>. <source>BMC bioinformatics</source> <volume>18</volume>, <fpage>78</fpage>&#x2013;<lpage>79</lpage>. <pub-id pub-id-type="doi">10.1186/s12859-017-1483-5</pub-id> </citation>
</ref>
<ref id="B10">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Ohno</surname>
<given-names>S.</given-names>
</name>
<name>
<surname>Shimizu</surname>
<given-names>H.</given-names>
</name>
<name>
<surname>Furusawa</surname>
<given-names>C.</given-names>
</name>
</person-group> (<year>2014</year>). <article-title>Fastpros: Screening of Reaction Knockout Strategies for Metabolic Engineering</article-title>. <source>Bioinformatics</source> <volume>30</volume>, <fpage>981</fpage>&#x2013;<lpage>987</lpage>. <pub-id pub-id-type="doi">10.1093/bioinformatics/btt672</pub-id> </citation>
</ref>
<ref id="B11">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Orth</surname>
<given-names>J.&#x20;D.</given-names>
</name>
<name>
<surname>Conrad</surname>
<given-names>T. M.</given-names>
</name>
<name>
<surname>Na</surname>
<given-names>J.</given-names>
</name>
<name>
<surname>Lerman</surname>
<given-names>J.&#x20;A.</given-names>
</name>
<name>
<surname>Nam</surname>
<given-names>H.</given-names>
</name>
<name>
<surname>Feist</surname>
<given-names>A. M.</given-names>
</name>
<etal/>
</person-group> (<year>2011</year>). <article-title>A Comprehensive Genome-Scale Reconstruction of <italic>Escherichia coli</italic> Metabolism--2011</article-title>. <source>Mol. Syst. Biol.</source> <volume>7</volume>, <fpage>535</fpage>. <pub-id pub-id-type="doi">10.1038/msb.2011.65</pub-id> </citation>
</ref>
<ref id="B12">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Patil</surname>
<given-names>K. R.</given-names>
</name>
<name>
<surname>Rocha</surname>
<given-names>I.</given-names>
</name>
<name>
<surname>F&#xf6;rster</surname>
<given-names>J.</given-names>
</name>
<name>
<surname>Nielsen</surname>
<given-names>J.</given-names>
</name>
</person-group> (<year>2005</year>). <article-title>Evolutionary Programming as a Platform for In Silico Metabolic Engineering</article-title>. <source>BMC bioinformatics</source> <volume>6</volume>, <fpage>308</fpage>. <pub-id pub-id-type="doi">10.1186/1471-2105-6-308</pub-id> </citation>
</ref>
<ref id="B13">
<citation citation-type="book">
<person-group person-group-type="author">
<name>
<surname>Stephanopoulos</surname>
<given-names>G.</given-names>
</name>
<name>
<surname>Aristidou</surname>
<given-names>A. A.</given-names>
</name>
<name>
<surname>Nielsen</surname>
<given-names>J.</given-names>
</name>
</person-group> (<year>1998</year>). <source>Metabolic Engineering: Principles and Methodologies</source>. <publisher-name>Cambridge University Press</publisher-name>.</citation>
</ref>
<ref id="B14">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Tamura</surname>
<given-names>T.</given-names>
</name>
</person-group> (<year>2018</year>). <article-title>Grid-based Computational Methods for the Design of Constraint-Based Parsimonious Chemical Reaction Networks to Simulate Metabolite Production: Gridprod</article-title>. <source>BMC bioinformatics</source> <volume>19</volume>, <fpage>325</fpage>. <pub-id pub-id-type="doi">10.1186/s12859-018-2352-6</pub-id> </citation>
</ref>
<ref id="B15">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Tamura</surname>
<given-names>T.</given-names>
</name>
</person-group> (<year>2021a</year>). <article-title>Efficient Reaction Deletion Algorithms for Redesign of Constraint-Based Metabolic Networks for Metabolite Production with Weak Coupling</article-title>. <source>IPSJ&#x20;Trans. Bioinformatics</source> <volume>14</volume>, <fpage>12</fpage>&#x2013;<lpage>21</lpage>. <pub-id pub-id-type="doi">10.2197/ipsjtbio.14.12</pub-id> </citation>
</ref>
<ref id="B16">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Tamura</surname>
<given-names>T.</given-names>
</name>
</person-group> (<year>2021b</year>). <article-title>L1 Norm Minimal Mode-Based Methods for Listing Reaction Network Designs for Metabolite Production</article-title>. <source>IEICE Trans. Inf. Syst.</source> <volume>E104.D</volume>, <fpage>679</fpage>&#x2013;<lpage>687</lpage>. <pub-id pub-id-type="doi">10.1587/transinf.2020edp7247</pub-id> </citation>
</ref>
<ref id="B17">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>von Kamp</surname>
<given-names>A.</given-names>
</name>
<name>
<surname>Klamt</surname>
<given-names>S.</given-names>
</name>
</person-group> (<year>2017</year>). <article-title>Growth-coupled Overproduction Is Feasible for Almost All Metabolites in Five Major Production Organisms</article-title>. <source>Nat. Commun.</source> <volume>8</volume>, <fpage>15956</fpage>. <pub-id pub-id-type="doi">10.1038/ncomms15956</pub-id> </citation>
</ref>
<ref id="B18">
<citation citation-type="journal">
<person-group person-group-type="author">
<name>
<surname>Yang</surname>
<given-names>L.</given-names>
</name>
<name>
<surname>Cluett</surname>
<given-names>W. R.</given-names>
</name>
<name>
<surname>Mahadevan</surname>
<given-names>R.</given-names>
</name>
</person-group> (<year>2011</year>). <article-title>Emilio: a Fast Algorithm for Genome-Scale Strain Design</article-title>. <source>Metab. Eng.</source> <volume>13</volume>, <fpage>272</fpage>&#x2013;<lpage>281</lpage>. <pub-id pub-id-type="doi">10.1016/j.ymben.2011.03.002</pub-id> </citation>
</ref>
</ref-list>
</back>
</article>