In PriceLab/trena: Fit transcriptional regulatory networks using gene expression, priors, machine learning

Introduction

The recently published Benchmark and integration of resources for the estimation of human transcription factor activities provides a useful curation of TF/target gene relations divided into five confidence categories. We chose one high confidence (score "A") relation from this dataset for this case study: the regulation of NFE2 by GATA1 in erythropoiesis. We will use trena to "discover" this relationship, a necessary but not sufficient proof of trena's value.

Hematopoiesis (the production of blood cells) involves two subprocesses: erythropoiesis (red blood cell biogenesis) and megakaryopoiesis (megakaryocytes and platelets). These processes have been intensively studied for decades, and much has been learned about the complex and hierarchical regulation of these processes of differentiation. GATA1, TALl, KLF1 and the p45 isoform of NFE2 are some of the transcription factors involved.

The Benchmark dataset reports two studies[ref1, ref2] reporting transcriptional regulation of NFE2 by GATA1. In addition, Takayama et al, 2010 report that "In megakaryocytes, GATA1 deficiency reduces p45 expression by approximately 50%, indicating the presence of GATA1-dependent and -independent regulation of the p45 gene." We lack precise knowledge of the intricacies of regulatory networks in erythropoieis, but the cited research in combination strongly suggests the regulation by GATA1 of NFE2 in erythropoieis.

Corces et al, 2016, RNA-seq on FACS sorted cells in erythropoiesis

PriceLab/trena documentation built on March 16, 2023, 10:01 a.m.