Reliable scaling of Position Weight Matrices for binding strength comparisons between transcription factors

TL;DR

This paper introduces two methods to accurately scale PWM scores, enabling direct comparison of transcription factor binding strengths across different PWMs and biological contexts.

Contribution

The authors present novel approaches to determine the scaling parameter λ for PWMs, facilitating reliable energy inference and cross-PWM comparisons for transcription factors.

Findings

λ distributions align with DNA binding properties across TF families

Methods produce consistent scaling parameters

Enables direct comparison of PWMs from different sources

Abstract

Scoring DNA sequences against Position Weight Matrices (PWMs) is a widely adopted method to identify putative transcription factor binding sites. While common bioinformatics tools produce scores that can reflect the binding strength between a specific transcription factor and the DNA, these scores are not directly comparable between different transcription factors. Here, we provide two different ways to find the scaling parameter $\lambda$ that allows us to infer binding energy from a PWM score. The first approach uses a PWM and background genomic sequence as input to estimate $\lambda$ for a specific transcription factor, which we applied to show that $\lambda$ distributions for different transcription factor families correspond with their DNA binding properties. Our second method can reliably convert $\lambda$ between different PWMs of the same transcription factor, which allows us to directly compare PWMs that were generated by different approaches. These two approaches provide consistent and computationally efficient ways to scale PWMs scores and estimate transcription factor binding sites strength.