GFSAT

GFSAT computes gene functional similarity using semantic similarity measurements derived from Gene Ontology (GO) annotations to quantify relationships among genes.

Key Features:

• Shortest Semantic Differentiation Distance (SSDD): A novel GO-intrinsic semantic similarity metric inspired by concepts of cellular differentiation and dedifferentiation.
• GO-intrinsic method: SSDD is intrinsic to Gene Ontology (GO) and does not rely on external datasets.
• Handling identical annotations: Distinguishes between identical GO annotations to improve accuracy of similarity assessments.
• Bias reduction: Mitigates bias toward well-annotated proteins by not depending on annotation richness.
• Comparison modes: Facilitates functional similarity comparisons for two genes, two gene groups, and pairwise gene groups.
• Validation and performance: Validated using human ratings and a benchmark dataset, demonstrating improved performance over existing semantic similarity methods.

Scientific Applications:

• Gene function prediction: Enhances prediction of gene functions using GO-based semantic similarity measures.
• Protein-protein interaction studies: Facilitates exploration of potential protein-protein interactions by comparing functional similarities.
• Disease research: Assists in identifying genes with similar functions that may be implicated in specific diseases.
• Functional genomics and systems biology: Supports analysis of gene function relationships in functional genomics and systems biology studies.

Methodology:

GFSAT employs the Shortest Semantic Differentiation Distance (SSDD), a GO-intrinsic semantic similarity metric inspired by cellular differentiation and dedifferentiation, avoids reliance on external datasets, and was validated using human ratings and a benchmark dataset.