Network-Driven Computational Framework Identifies FDA-Approved Drug Repurposing Across Heterogeneous Brain Cancers

Om Prakash^*

• The Institute of Mathematical Sciences, Chennai, India

Background: Brain cancers are notorious for their heterogeneity, which complicates therapeutic decision, because of its recurrently dysregulated signaling pathways. Cancer system characteristics has identified key components involved in brain cancer, such as EGFR, BRAF, PDGFRA, TP53, MGMT, CDK1/2/3/4, COX-1/2, VEGFR2, TERT, and CYP2D6 along with U87 cell line. These components are the core attention for designing of protocols for rational drug selection. For drug repurposing, designed protocols are generally hypothesized with 'Food and Drug Administration' (FDA)-approved drugs. Methodology: In present study, a protocol was designed to address this complexity using the identified pathway components. These components served as the basis for defining the signature of small molecules. The set of molecular signatures utilized to develop a network-driven computational framework. Two in-house applications were developed here: first, 'in-mac', a molecular profile-generator; and second, a network-based database (ReBrain) derived from FDA-approved drug molecules. In-mac is a computational bioassay platform for generating activity profile of small molecule, while 'ReBrain' is a database for broad-spectrum drug-repurposing analytics. Performance of profile-set was evaluated and validated with five machine learning models with three different classified datasets. Results: A total of 2,809 FDA-approved drug-molecules (molecular weight ≤ 500 Dalton) were profiled using in-mac. Each individual molecular profile included fifteen-dimensional activity signatures. Profile-set was proved to be significantly potential for drug-repurposing. These molecular profiles were used to perform regression analysis, followed by calculation of inter-molecular Euclidian distances and development of an inter-molecule network. The ReBrain platform also enabled in silico knockout or knock-in facility of specific pathway components. Network refinement was facilitated based on molecular weight and distance thresholds. Conclusion: The proposed profile-network-based method achieved 70% to 95% accuracy for drug repurposing across different disease categories related with brain. In-mac and ReBrain were used for drug repurposing of known drugs against brain cancer. As a result, three repurposed drugs were identified as priorities: (i) Mefloquine (reference drug: Vorasidenib Citrate), (ii) Clofibric Acid (reference drug: Carmustine), and Armillarisin A (reference drug: Lomustine). These results also suggest the repurposing candidates for synergistic combinations also across different brain tumors. Applications are freely accessible in public domain at (https://assay.smallmoles.com/escorwin).