Link to bioRxiv paper:
http://biorxiv.org/cgi/content/short/2022.10.17.512527v1?rss=1
Authors: Wang, J., Wan, Y.-W., Al-Ouran, R., Liu, Z.
Abstract:
Scientists have learned much about gene expression in the past few years. Integrating large-scale gene-expression data will significantly bolster researchers' understanding of biology and diseases. While we currently use co-expression models to discover gene-gene associations in RNA seq data, integrating RNA seq data generated from different experiments can have a "batch effect", which decreases data quality and makes it harder to glean definitive gene relationships from co-expression models. The biological relationship can also be context-dependent and non-linear, which make them difficult to detect using current co-expression approaches. Here, we propose a co-perturbation model that better identifies gene-gene associations in integrated data and reveals non-linear correlation between genes.
Copy rights belong to original authors. Visit the link for more info
Podcast created by Paper Player, LLC
