This data repository contains coexpression networks from publicly-available RNA-Seq datasets (obtained from the recount2 database) that were generated using the best workflows identified in the benchmarking study: Johnson KA, Krishnan A (2020) Robust normalization and transformation techniques for constructing gene coexpression networks from RNA-seq data. bioRxiv 10.1101/2020.09.22.308577. GTEx coexpression networks There are 62 coexpression networks built from 31 GTEx datasets (each dataset corresponding to one GTEx tissue) reconstructed using two different network-building workflows: i) CTF_CLR: Counts adjusted using TMM Factors followed by CLR transformation of the Pearson correlation coefficients; ii) CTF: Counts adjusted using TMM Factors (without any further transformation). SRA coexpression networks There are 256 coexpression networks built from 256 SRA datasets. Each dataset corresponds to a set of samples generated as part of the same transcriptome experiment from the same tissue. These networks are reconstructed using the top-performing workflow: CTF, Counts adjusted using TMM Factors. Refer to the preprint for more details on the workflows and the steps used for obtaining the original datasets.

This work was primarily supported by US National Institutes of Health (NIH) grants R35 GM128765 to A.K. and supported in part by MSU start-up funds to A.K.