ICoVeR – Interactive Contig-bin Verification and Refinement tool

Context

Advances in high-throughput sequencing now allow for much deeper exploitation of natural and engineered microbial communities, and make it possible to unravel so called “microbial dark matter” (microbes that until now have evaded cultivation). Metagenomics results in a large number of genomic fragments (contigs) that need to be grouped (binned) in order to reconstruct draft genomes of novel bacteria. Even though many novel contig binning algorithms have appeared recently, they often lack consensus and result in different bin assignments. Therefore, further verification of bin completeness and contamination and bin refinement is necessary.

Overview

ICoVeR is an interactive framework that allows the visualization of several binning outputs in order to perform further supervised (human-assisted) improvement of automatically generated genome bins. Its interactive visualization interface is composed of two connected main views:

1) the parallel coordinates view in which GC content, contig length, sample coverages as well as tetra nucleotide frequencies (TNFs) are displayed

2) the dimensionality reduction plot in which projections of the TNFs are shown.

ICoVeR is an R plugin and as such is easily deployable on any system for which R is available. Data management and analytical functions are implemented in R. The frontend of the application has been developed using web technologies. The parallel coordinates visualization technique allows the user to display a large number of variables at the same time. On a standard 19” screen with a resolution of 1280x1024, 35 variables can be easily displayed, while still enabling the user to search for patterns in a meaningful way (easily scalable to large datasets).

Benefits

• visual interface enables verification and refinement of genome bins assigned by multiple automated algorithms, leading to nearly complete microbial draft genomes
• easy to install and developed as a plugin for the R language, thus it integrates well with the tool chains commonly used by microbiologists
• scales well to large metagenomic datasets, both in terms of the number of samples as well as bin assignments

Potential Applications

The reconstruction of microbial genomes from metagenomic reads helps to unravel the genetic potential of novel, usually not yet cultivable microbes from both natural and engineered environments. Therefore, among others, it is of major interest to:

• food fermentation processes (lactic fermentation, food and feed bio-preservation, etc.)
• soil studies (e.g. agriculture and bioremediation purposes)
• bioenergy sector and green chemistry(e.g. anaerobic digestion, carboxylation platform, biohydrogen production)
• human gut and associated disorders related to microflora perturbation (e.g. diabetes, Crohn's disease, asthma, autism, etc.)
• any other microbe-driven process of industrial interest