TiSAn

TiSAn predicts tissue-specific functional effects of coding and non-coding genetic variants to prioritize variants relevant to particular tissues for interpreting complex traits and diseases.

Key Features:

• Integration of Genome-Scale Data: Integrates cross-tissue epigenomic data with multiple genome-scale datasets to provide comprehensive functional annotations.
• Machine Learning Approach: Uses advanced machine learning techniques to discriminate tissue-relevant from non-relevant variants on a genome-wide scale.
• Predictive Models for Specific Tissues: Provides predictive models developed for human heart and brain tissues, exemplified by TiSAn-heart and TiSAn-brain linked to coronary artery disease and autism spectrum disorder respectively.
• Enhanced Variant Prioritization: Improves prioritization of genetic variants relative to existing methods such as GenoSkyLine, aiding filtering for whole-genome sequencing and genome-wide association studies.

Scientific Applications:

• Interpretation of Complex Traits and Diseases: Assists interpretation of complex traits and diseases by predicting tissue-specific impacts of deleterious variants.
• Disease-Specific Variant Discovery: Applied to identify tissue-relevant variants associated with autism spectrum disorder (TiSAn-brain) and coronary artery disease (TiSAn-heart).
• Variant Prioritization for WGS and GWAS: Supports variant prioritization in whole-genome sequencing and genome-wide association studies to inform potential therapeutic targets and personalized medicine.

Methodology:

Integrates cross-tissue epigenomic data with additional genome-scale datasets specific to the tissue of interest and applies machine learning models trained to distinguish tissue-relevant variants.