An approach to identify over-represented cis-elements in related sequences

Area of Science:

• Computational biology
• Bioinformatics
• Genomics

Background:

• Transcription factor binding site (TFBS) identification is crucial in computational biology.
• Positional weight matrices (PWMs) model TFBS sequence patterns but often require arbitrary user-defined thresholds.
• Existing methods suffer from low specificity, hindering accurate interpretation of TFBS prediction results.

Purpose of the Study:

• To develop a computational method for identifying over-represented cis-elements using PWM-based similarity scores.
• To overcome limitations of user-defined thresholds in TFBS prediction.
• To enhance the specificity and interpretability of TFBS identification.

Main Methods:

• Developed a novel method utilizing PWM-based similarity scores to identify cis-elements.
• Analyzed three sets of closely related promoters.
• Implemented automatic threshold determination for similarity scores.

Main Results:

• Successfully identified over-represented motifs with high PWM similarity scores.
• Demonstrated automatic threshold determination for similarity scores.
• Provided an online program for TFBS identification.

Conclusions:

• The developed method improves the identification of transcription factor binding sites.
• Automated threshold determination enhances the reliability and interpretability of TFBS predictions.
• The approach offers a valuable tool for further research in cis-element analysis.