Simultaneous alignment and annotation of cis-regulatory regions

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Current methods for identifying conserved transcription factor binding sites in regulatory regions perform alignment and annotation separately.
• This sequential approach faces challenges with weak site descriptions or low sequence similarity due to alignment's local gap structure.
• A simultaneous approach is desirable to effectively integrate alignment and potential site location detection.

Purpose of the Study:

• To develop a novel tool, SimAnn, that simultaneously aligns sequences and annotates transcription factor binding sites.
• To improve the detection of conserved binding sites by integrating alignment and annotation steps.
• To provide a statistically grounded method for parameter selection applicable to various binding site models.

Main Methods:

• Developed SimAnn, a unified algorithm combining sequence alignment and binding site annotation.
• Implemented a statistical approach for calculating input parameters.
• Evaluated SimAnn's performance against non-simultaneous methods using simulated and real biological data.

Main Results:

• SimAnn successfully detects conserved transcription factor binding sites more clearly by integrating alignment and annotation.
• The tool demonstrates robust performance across simulated and real datasets, outperforming existing methods.
• Statistically derived parameters enable SimAnn's flexible application to diverse binding site models.

Conclusions:

• SimAnn offers a significant advancement in identifying conserved transcription factor binding sites.
• The simultaneous alignment and annotation approach overcomes limitations of traditional methods.
• The tool is available as a C++ implementation with accompanying Perl scripts for parameter calculation.