T-Gene: improved target gene prediction.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Identifying transcription factor (TF) target genes is crucial for understanding gene regulation.
• Previous methods like CisMapper relied on histone modification and gene expression data, limiting their applicability.
• These methods did not explicitly consider the genomic distance between TF binding sites and target genes.

Purpose of the Study:

• To develop a novel algorithm, T-Gene, for predicting transcription factor target genes.
• To overcome the limitations of previous methods by incorporating genomic distance and histone/expression correlation.
• To provide a user-friendly tool for TF target gene prediction.

Main Methods:

• The T-Gene algorithm calculates a novel score integrating genomic distance and histone/expression correlation.
• The algorithm predicts which genes are likely regulated by a TF and which binding sites are involved.
• T-Gene can also predict targets using distance alone when extensive data is unavailable.

Main Results:

• T-Gene achieved a median precision above 60% in predicting regulatory element-promoter contact.
• The algorithm demonstrated median precision above 40% using distance alone for organisms with limited data.
• T-Gene provides statistical significance estimates for its predictions.

Conclusions:

• T-Gene offers an improved and versatile approach for identifying transcription factor target genes.
• The algorithm enhances the understanding of gene regulation by integrating multiple predictive factors.
• T-Gene is accessible via a web server and command-line tool, facilitating broader research applications.