B.Tech project on functionally redundant genome clusters

Abstract

This project aims to develop computational pipelines to cluster microbiomes from different geographic locations and disease phenotype based on the similarity in their functional signatures. This initiative has built an extensive database of functional feature count matrices for 71,140 high-quality metagenomes from their annotation files. The database will give information about several functional features like the Carbohydrate active enzymes for different microbial species, Orthologous groups, KEGG reaction pathways, etc. Locality-sensitive hashing techniques have been deployed to find the cluster of functionally similar genomes/taxa as well as similar microbiomes based on the functional profiles. This framework will be implemented to develop better diagnostic methodologies for the human gut microbiome.