Abstract
This dissertation centers on the role of methodological refinement and assessment in population genomics and metagenomics, focusing on two key aspects: (1) statistical estimation of parameters in large evolutionary models using Approximate BayesianComputation (ABC); (2) evaluation of precision in high-throughput chromosome conformation capture (Hi-C) technology for robust spatial insights into microbiome.
In the first project, computational challenges of deriving exact model likelihoods are addressed by leveraging mechanistically motivated forward-time simulation model and employing ABC. The developed model incorporates divergent selection, variable migration rates, modes of reproduction (sexual/asexual), length and number of migration-selection cycles. The research assesses the computational feasibility of ABC and evaluates the quality of estimates regarding loci under selection andselection strength. The impact of genetic drift and recombination rate on estimation are assessed.
The second project assesses Hi-C’s precision in recovering metagenome-assembled genomes (MAGs) through ProxiMeta, Bin3C, and MetaTOR pipelines. The project employs precision metrics, statistical testing, and assessment of algorithmic choicesand Hi-C libraries, in obtaining dependable results, contributing to understanding the influence of microbiome analysis methodologies.
Together, these projects emphasize the significance of methodological advancement in population genomics and metagenomics, contributing to the development of improved analytical tools.