AUTHOR=Pennells Jordan , Martin Darren J. 

TITLE=Statistical genetics concepts in biomass-based materials engineering

JOURNAL=Frontiers in Bioengineering and Biotechnology

VOLUME=Volume 10 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/bioengineering-and-biotechnology/articles/10.3389/fbioe.2022.1022948

DOI=10.3389/fbioe.2022.1022948

ISSN=2296-4185

ABSTRACT=With the rise of biomass-based materials such as nanocellulose, there is a growing need to develop statistical methods capable of leveraging inter-dependent experimental data to improve material design, product development, and process optimisation. Statistical approaches are essential given the multifaceted nature of variability in lignocellulosic biomass, which includes a range of different biomass feedstock types, a combinative arrangement of biomass processing routes, and an array of different product formats depending on the focal application. To account for this variability and to extract meaningful patterns from research studies, there is a requirement to generate biomass-derived material property data through well-designed experimental systems that enable statistical analysis. To drive this trend, this article proposes the cross-disciplinary utilisation of statistical modelling approaches commonly applied within the field of statistical genetics to evaluate data generated in biomass-based material research. The concepts of variance partitioning, heritability, hierarchical clustering, and selection gradients have been explained in their native context of statistical genetics and subsequently applied across the disciplinary boundary to evaluate relationships within a model experimental study involving the production of sorghum-derived cellulose nanofibres and their subsequent fabrication into nanopaper material. Variance partitioning and heritability calculates the relative influence of biomass vs processing factors on material performance, while hierarchical clustering highlights the similarity between experimental samples or characterisation metrics, and selection gradients elucidate the relationships between characterisation metrics and material quality. Ultimately, these statistical modelling approaches provide more depth to biomass-based material evaluation and outline a framework for robust product quality evaluation in nanocellulose production.