Fig. 1From: PredicTF: prediction of bacterial transcription factors in complex microbial communities using deep learningPredicTF workflow and testing. We collected publicly available data on TFs from two different databases: CollecTF and UniProtKB. After removing redundancies and filtering TFs well characterized, this data (BacTFDB) was used to train a deep learning model to predict new TFs and their families. Five model organisms (Escherichia coli, Bacillus subtillis, Pseudomonas fluorescens, Azotobacter vinelandii and Caulobacter crescentus) were used to test the accuracy of PredicTF. Later, we used the same approach to predict TFs from an isolate (P. aeruginosa) and mapped TFs predicted in transcriptomics data (P. aeruginosa and mutants in two experimental conditions). Finally, we used our tool to predict TF in complex communities (metagenome) and mapped these TFs in their respective meta-transcriptomesBack to article page