Advancing Multi-Locus Epistasis Detection Through Modified GAMETES Tool

Kalpana, Das (2023) Advancing Multi-Locus Epistasis Detection Through Modified GAMETES Tool. Masters thesis, Indian Institute of Science Education and Research Kolkata.

Text (MS dissertation of Kalpana Das (18MS187)) Thesis_18MS187.pdf - Submitted Version Restricted to Repository staff only Download (1MB)

Abstract

Understanding the underlying complex multi-locus interactions is essential for characterizing the true genetic basis for multifactorial human diseases. This requires unraveling the epistatic effects among the disease-contributing genes. Although several algorithms exist for generating synthetic multi-locus models through simulations, most are limited to two-locus models. Others not bound by order of epistasis are computationally expensive, inflexible, and unable to produce “gold standard” pure and strict epistatic models, which present the worst possible scenario for epistasis detection. Leveraging the algorithmic framework of the open-source GAMETES simulator, known for its ability to produce complex biallelic SNP models, we revisit the basic concepts and terminologies in epistasis detection. We explain each component of model generation and the intuition behind the mathematics of penetrance tables. We propose two modified versions of GAMETES, gametesdouble and gametes5, and compare the summaries of their model generation frequencies against that of GAMETES. We verify that both gametesdouble and gametes5 rapidly and precisely generate random, pure and strict epistatic models over different sets of model constraints. We also observe that they outperform GAMETES in generating models with higher orders of interactions and higher heritability values. The modified versions are freely available for download at https://github.com/KalpanaDas/gametesdouble and https://github.com/KalpanaDas/gametes5.

Actions (login required)

View Item