MoRAine--a web server for fast computational transcription factor binding motif re-annotation

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Accurate identification of transcription factor binding motifs (TFBMs) is experimentally challenging and time-consuming.
• Current TFBM annotations in databases often have mixed orientations, reducing the information content of position frequency matrices (PFMs) and sequence logos.
• This reduced information content impacts the accuracy of predicting further TFBMs.

Purpose of the Study:

• To develop an automated method for re-annotating TFBMs to improve PFM information content.
• To enhance the performance of TFBM prediction and classification using re-annotated motifs.
• To assess the impact of re-annotation on PFM quality and predictive accuracy.

Main Methods:

• Developed MoRAine, an algorithm that re-annotates TFBMs by comparing motifs and optimizing strand orientation and position.
• Employed two heuristic strategies to maximize the information content of adjusted PFMs.
• Evaluated MoRAine's performance using specificity, sensitivity, true positive, and false positive rates for TFBM predictions in E. coli.

Main Results:

• MoRAine significantly improves the information content of sequence logos derived from PFMs.
• The algorithm enhances the classification performance of PFM-based TFBM predictions.
• Re-annotated motifs using MoRAine lead to considerably increased prediction accuracy.

Conclusions:

• MoRAine offers an effective preprocessing step to improve TFBM identification and prediction.
• The algorithm is available as a publicly accessible web server and a downloadable stand-alone version.
• MoRAine facilitates more accurate genomic analysis by optimizing TFBM annotations.