SVD-Phy

SVD-Phy applies truncated singular value decomposition to phylogenetic profiles to predict functional associations between non-homologous genes by comparing their phylogenetic distributions.

Key Features:

• Truncated Singular Value Decomposition: Employs truncated SVD to reduce noise and enhance signal in phylogenetic profiles.
• Prediction of Functional Associations: Predicts functional associations between non-homologous genes by comparing their phylogenetic distributions.
• Reduction of False Positives: Mitigates the impact of uninformative profiles to decrease false positive association predictions.
• Benchmarking: Performance has been benchmarked against the KEGG pathway database to evaluate prediction accuracy.

Scientific Applications:

• Functional Genomics: Infers gene function and aids discovery of novel biological pathways and interactions not evident from sequence homology alone.
• Systems Biology: Supports reconstruction of functional interaction networks and exploration of regulatory relationships.
• Evolutionary Analysis: Assesses evolutionary relationships through comparative phylogenetic profiling of genes.

Methodology:

Comparison of gene phylogenetic distributions followed by truncated singular value decomposition on phylogenetic profiles; benchmarking against the KEGG pathway database.