PTF-Vāc:Ab-initiodiscovery of plant transcription factors binding sites using deep co-learning encoders-decoders

Abstract

AbstractDiscovery of transcription factors (TFs) binding sites (TFBS) and their motifs in plants pose significant challenges due to high cross-species variability. The interaction between TFs and their binding sites is highly specific and context dependent. Most of the existing TFBS finding software tools are not accurate enough to discover these binding sites in plants. They fail to capture the cross-species variability, interdependence between TF structure and its TFBS, and context specificity of binding. Since they are coupled to predefined TF specific model/matrix, they are highly vulnerable towards the volume and quality of data provided to build the motifs. All these software make a presumption or expect that the user input would be specific to any particular TF which renders them of very limited uses. This all makes them hardly of any use for purposes like genomic annotations of newly sequenced species. Here, we report a Deep Encoders-Decoders system, PTF-Vāc, founded on a universal model of deep co-learning on variability in binding sites and TF structure, making it completely free from the bottlenecks mentioned above. It has successfully decoupled the process of TFBS discovery from the prior step of motif finding and requirement of TF specific motif models. Due to a universal model for TF:DNA interactions as its guide, it can discover the binding sites in total independence from data volume, species and TF specific models. It can accurately detect even the binding sites of never seen before TFs families and species and can be used to define credible motifs from its TFBS report.