Machine Learning Uncovers a Data-Driven Transcriptional Regulatory Network for the Crenarchaeal Thermoacidophile Sulfolobus acidocaldarius

Abstract

Dynamic cellular responses to environmental constraints are coordinated by the transcriptional regulatory network (TRN), which modulates gene expression. This network controls most fundamental cellular responses, including metabolism, motility, and stress responses. Here, we apply independent component analysis, an unsupervised machine learning approach, to 95 high-quality Sulfolobus acidocaldarius RNA-seq datasets and extract 45 independently modulated gene sets, or iModulons. Together, these iModulons contain 755 genes (32% of the genes identified on the genome) and explain over 70% of the variance in the expression compendium. We show that five modules represent the effects of known transcriptional regulators, and hypothesize that most of the remaining modules represent the effects of uncharacterized regulators. Further analysis of these gene sets results in: (1) the prediction of a DNA export system composed of five uncharacterized genes, (2) expansion of the LysM regulon, and (3) evidence for an as-yet-undiscovered global regulon. Our approach allows for a mechanistic, systems-level elucidation of an extremophile’s responses to biological perturbations, which could inform research on gene-regulator interactions and facilitate regulator discovery in S. acidocaldarius. We also provide the first global TRN for S. acidocaldarius. Collectively, these results provide a roadmap toward regulatory network discovery in archaea.

A multi-scale transcriptional regulatory network knowledge base forEscherichia coli

“…Taken together, PRECISE-1K and iModulons extracted from it highlight the central role that top-down, data-driven methods must take in transcriptional regulatory network discovery across organisms. Indeed, iModulons have already successfully generated top-down regulatory networks for other organisms (Chauhan et al, 2021; Lim et al, 2022; Poudel et al, 2020; Rajput et al, 2022; Rychel et al, 2020; Sastry et al, 2019; Yoo et al, 2022). The success of PRECISE-1K serves to further cement both the importance of pursuing such efforts and the reliability of the results.…”

“…Independent component analysis (ICA) (Comon, 1994) is a signal processing algorithm that outperforms other methods for the extraction of biologically meaningful regulatory modules from gene expression data (Saelens et al, 2018). Application of this method to publicly-available prokaryotic expression data has consistently recovered TRN modules across organisms (Chauhan et al, 2021; Poudel et al, 2020; Rajput et al, 2022; Rychel et al, 2020; Sastry et al, 2019; Yoo et al, 2022). ICA’s effectiveness results from its ability to identify independent groups of genes that vary consistently across samples, regardless of group size or overlapping membership.…”

Machine Learning of All Mycobacterium tuberculosis H37Rv RNA-seq Data Reveals a Structured Interplay between Metabolism, Stress Response, and Infection

“…Additionally, we found both the WhiB1 and GroEL/ES complex iModulons play a role in protein synthesis. WhiB1 also contains several genes that code for RNA polymerase subunits, and is likely a translation iModulon that has been seen in the ICA decompositions of other organisms ( 9 , 10 , 45 ). All three iModulons are related to growth and replication, which suggests that cell division is an important response in M. tuberculosis in a decreasing oxygen environment.…”

Coordination of CcpA and CodY regulators in Staphylococcus aureus USA300 strains

“…and genes involved in translation such as infA and fusA which encode translation initiation factor IF-1 and elongation factor G respectively(Figure 2b). This iModulon has been enriched in almost all bacteria and archaea for which iModulons have been calculated 20,[22][23][24][25] .…”