Improvement of TRANSFAC matrices using multiple local alignment of transcription factor binding site sequences

Y Fu; Z Weng

doi:10.1109/IEMBS.2004.1403814

Improvement of TRANSFAC matrices using multiple local alignment of transcription factor binding site sequences

Conf Proc IEEE Eng Med Biol Soc. 2004:2004:2856-9. doi: 10.1109/IEMBS.2004.1403814.

Authors

Y Fu¹, Z Weng

Affiliation

¹ Bioinformatics Program, Boston Univ., MA, USA.

PMID: 17270873
DOI: 10.1109/IEMBS.2004.1403814

Abstract

This paper describes a novel approach to constructing position-specific weight matrices (PWM) based on the transcription factor binding site (TFBS) data provide by the TRANSFAC database, and the comparison of the thus generated PWMs with the original TRANSFAC matrices. Multiple local sequence alignment was performed on the TFBSs of each transcription factor. Several different alignment programs were tested and their matrices were compared to the original TRANSFAC matrices. One of the alignment programs, GLAM, produced comparable matrices in terms of the average ranking of true positive sites across the whole test set of sequences.