Back

Predicting input signals of transcription factors in Escherichia coli through metabolomics and transcriptomics

Trouillon, J.; Huber, A. E.; Trabesinger, Y.; Sauer, U.

2025-02-24 microbiology
10.1101/2025.02.24.639830 bioRxiv
Show abstract

The activity of bacterial transcription factors (TFs) is typically modulated through direct interactions with small molecules. However, these input signals remain unknown for most TFs, even in well-studied model bacteria. Identifying these signals typically requires tedious experiments for each TF. Here, we develop a systematic workflow for the identification of TF input signals in bacteria based on metabolomics and transcriptomics data. We inferred the activity of 173 TFs from published transcriptomics data and determined the abundance of 279 metabolites across 40 matched experimental conditions in Escherichia coli. By correlating TF activities with metabolites abundances, we successfully identified previously known TF-metabolite interactions and predicted novel TF effector metabolites for 41 TFs. To validate our predictions, we conducted in vitro assays and confirmed a predicted effector metabolite for LeuO. As a result, we established a network of 80 regulatory interactions between 71 metabolites and 41 E. coli TFs. This network includes 76 novel interactions that encompass a diverse range of chemical classes and regulatory patterns, bringing us closer to a comprehensive TF regulatory network in E. coli.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.