MATLIGN: a motif clustering, comparison and matching tool

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Transcription factor binding sites (TFBS) are commonly represented as position frequency matrices (PFM) or consensus sequences (CS).
• Computational methods predict potential TFBS motifs, but post-processing is needed to match, compare, and cluster these predictions.
• Efficient post-processing reduces redundancy and aids in annotating predicted motifs.

Purpose of the Study:

• To develop a tool for post-processing DNA sequence motifs in PFM and CS formats.
• To facilitate the matching, comparison, and clustering of predicted and known TFBS motifs.
• To simplify the analysis of large collections of DNA sequence motifs.

Main Methods:

• Developed Matlign, a tool for aligning and evaluating motif similarity.
• Utilized a combination of scoring functions and hierarchical clustering for motif visualization.
• Employed an alignment algorithm that limits distinct gaps to correctly align motifs with internal spacers.
• Selected non-redundant motif sets by cutting hierarchical trees using silhouette values.

Main Results:

• Matlign reliably aligns and evaluates similarity between motifs.
• The tool correctly aligns motifs containing internal spacers.
• Matlign identifies the most similar analogue from a collection of characterized regulatory elements.
• The method aids in annotating motif predictions through PFM library searches.

Conclusions:

• Matlign is a user-friendly tool for post-processing large collections of DNA sequence motifs.
• It provides researchers with a non-redundant set of motifs from numerous potential regulatory motifs.
• The tool facilitates the association of predicted motifs with known regulatory elements.
• A web-server for Matlign is available for public use.