A complete framework for performing TF co-occurence and TF grammar analysis

TF-COMB

TF-COMB (Transcription Factor Co-Occurrence using Market Basket analysis) is a framework of methods for investigating transcription factor co-occurrence and and grammar of TF binding.

Introduction

In many cases, combinations of multiple TFs are needed to elicit a specific cellular response - a concept known as TF co-occurrence. In order to detect co-occurring TF binding sites, we developed TF-COMB (Transcription Factor Co-Occurrence using Market Basket analysis), which utilizes a flexible input of ChIP-seq peaks, motif positions, footprint locations, ATAC-seq peaks etc. to identify highly co-occurring TFs/regions. The association is calculated using an association method known as a market basket analysis. Market basket analysis has classically been applied to investigate shopping habits such as “if the customer buys cereal, they are likely to buy milk”, however, this approach can be applied to TF co-occurrence analysis like “if TF1 binds, it is also likely that TF2 binds” as well. This co-occurrence analysis is extended in the context of binding grammar, which is defined by the need for a syntax of TFBS binding, such as a certain arrangement, order, distance, affinity and/or relative orientation of the given TFBS.

overview of TF-COMB functionalities

Main features

The main features of TF-COMB are:

  • Easy to use
  • Global TF co-occurrence analysis
  • Preferred distance between TFs
  • Orientation specificity of stranded regions
  • Differential co-occurrence between conditions
  • Network analysis and visualization to identify protein hubs

TF-COMB can be used as a framework as illustrated below: