Machine learning uncovers independently regulated modules in the Bacillus subtilis transcriptome

Abstract

The transcriptional regulatory network (TRN) of Bacillus subtilis coordinates cellular functions of fundamental interest, including metabolism, biofilm formation, and sporulation. Here, we use unsupervised machine learning to modularize the transcriptome and quantitatively describe regulatory activity under diverse conditions, creating an unbiased summary of gene expression. We obtain 83 independently modulated gene sets that explain most of the variance in expression and demonstrate that 76% of them represent the effects of known regulators. The TRN structure and its condition-dependent activity uncover putative or recently discovered roles for at least five regulons, such as a relationship between histidine utilization and quorum sensing. The TRN also facilitates quantification of population-level sporulation states. As this TRN covers the majority of the transcriptome and concisely characterizes the global expression state, it could inform research on nearly every aspect of transcriptional regulation in B. subtilis.

Independent component analysis recovers consistent regulatory signals from disparate datasets

“…aureus and B . subtilis using single-source expression datasets [ 29 , 30 ]. In addition, ICA has been applied to human transcriptomic datasets to identify co-regulated gene sets [ 18 , 21 ], but characterization of many components was hindered due to the high fraction of unknown human genes relative to model bacteria [ 31 , 32 ].…”

Machine Learning of Pseudomonas aeruginosa transcriptomes identifies independently modulated sets of genes associated with known transcriptional regulators

“…Independent component analysis (ICA), a method to identify independent signals in complex data sets 5 , has been applied to data sets of bacterial transcriptomes to identify independently modulated sets of genes, called iModulons, and the transcriptional regulators that control them [6][7][8] . iModulons have been used to study the adaptive evolution trade-off during oxidative stress under naphthoquinone-based aerobic respiration 9 , mutations in the OxyR transcription factor and regulation of the ROS response 10 , and the host response to expression of heterologous proteins 11 .…”

“…iModulons have been used to study the adaptive evolution trade-off during oxidative stress under naphthoquinone-based aerobic respiration 9 , mutations in the OxyR transcription factor and regulation of the ROS response 10 , and the host response to expression of heterologous proteins 11 . We have also used ICA to elucidate the TRN structures of Escherichia coli 7 , Staphylococcus aureus 6 , and Bacillus subtilis 8 , which are presented in interactive dashboards on the iModulonDB.org website 12 .…”

Machine learning uncovers a data-driven transcriptional regulatory network for the Crenarchaeal thermoacidophile Sulfolobus acidocaldarius

“…Given the considerable insights provided by ICA, along with a proven track-record (Sastry et al, 2019;Poudel et al, 2020;Rychel et al, 2020), we applied it to a compendium of 95 publically available RNA-seq expression profiles for S. acidocaldarius to deconvolute its TRN. This is the first application of ICA towards deconvolution of an archaeal TRN, and has led towards the generation of the most complete, global TRN of S. acidocaldarius.…”

“…Despite the difference in approach (i.e. data analytics vs direct molecular methods), iModulons recapitulate many known regulons from the literature, and have accurately predicted new genetic targets for regulators (Sastry et al, 2019;Poudel et al, 2020;Rychel et al, 2020) and elucidated gene functions (Rodionova et al, 2020(Rodionova et al, , 2021. This top-down approach allows for an unbiased method for reconstructing the TRN of prokaryotic organisms of interest, including archaea.…”