SMASHing regulatory sites in DNA by human-mouse sequence comparisons

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Regulatory sequence elements are crucial for understanding gene expression.
• Systematic annotation of transcription factor binding sites in the human genome is lacking.
• Cross-species sequence comparisons are vital for identifying conserved regulatory elements.

Purpose of the Study:

• To develop a computational pipeline (SMASH) for annotating transcription factor binding sites in the human genome.
• To leverage human and mouse genome drafts for identifying conserved regulatory elements.
• To integrate predicted regulatory information with existing human genome annotations.

Main Methods:

• Developed SMASH, a computational pipeline for identifying orthologous genes and comparing upstream genomic regions between human and mouse.
• Utilized quasi-full length cDNA sequences to estimate transcription start sites.
• Employed a novel probabilistic method to identify conserved transcription factor binding sites, accounting for factor competition.
• Presented results via a genome browser web interface.

Main Results:

• Identified thousands of orthologous human/mouse proteins and mapped them to genomic sequences.
• Annotated putative regulatory elements in conserved, non-coding, upstream regions for approximately 2,500 gene pairs.
• SMASH's probabilistic method successfully identified conserved binding sites.
• Predicted regulatory information integrated with current human genome annotations.

Conclusions:

• SMASH provides a robust computational approach for annotating transcription factor binding sites and regulatory elements in the human genome.
• The pipeline effectively utilizes cross-species comparisons to identify conserved non-coding regulatory regions.
• SMASH results demonstrate favorable comparison to existing computational methods and are validated by experimental data.