A Scalable Reference-Free Metagenomic Binning Pipeline

摘要

Metagenomics studies microbial genomes in an ecosystem such as the gastrointestinal tract of a human through sequencing thousands of organism in parallel. The sheer number of genomic fragments are challenging for current metagenomic binning software to process. Here we present a scalable reference-free metagenomic binning pipeline designed to handle large scale metagenomic data. It allows users to input several tera base pairs (TB) of reads and produces highly accurate binning results, even at a species level. The pipeline outputs all binned species in multiple metagenomic samples and their estimated relative abundance. We integrate the pipeline into an open-source software, MetaMat.