VFDB 2016: hierarchical and refined dataset for big data analysis—10 years on

TLDR

The VFDB is a comprehensive, up‑to‑date repository of bacterial virulence factors that has served the field for over a decade, but the exponential growth of biological data now challenges its capacity for big‑data analysis. The authors removed redundancy, created a core dataset of experimentally verified VFs and a full dataset of all known and predicted VFs, and refined gene annotations using controlled vocabularies. These improvements enhanced data quality and increased the database’s usability for bioinformatic mining of the rapidly expanding bacterial VF data.

Abstract

The virulence factor database (VFDB, http://www.mgc.ac.cn/VFs/) is dedicated to providing up-to-date knowledge of virulence factors (VFs) of various bacterial pathogens. Since its inception the VFDB has served as a comprehensive repository of bacterial VFs for over a decade. The exponential growth in the amount of biological data is challenging to the current database in regard to big data analysis. We recently improved two aspects of the infrastructural dataset of VFDB: (i) removed the redundancy introduced by previous releases and generated two hierarchical datasets – one core dataset of experimentally verified VFs only and another full dataset including all known and predicted VFs and (ii) refined the gene annotation of the core dataset with controlled vocabularies. Our efforts enhanced the data quality of the VFDB and promoted the usability of the database in the big data era for the bioinformatic mining of the explosively growing data regarding bacterial VFs.