Sub-GOFA: A tool for Sub-Gene Ontology function analysis in clonal mosaicism using semantic (logical) similarity

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Clonal mosaicism, common in cancer patients, involves post-zygotic mutations in cellular subpopulations.
• Current analysis methods bundle mosaicism into large regions, overlooking low-frequency genes crucial for distinguishing patients.
• Identifying these subtle genetic differences is key to understanding cancer development and patient stratification.

Purpose of the Study:

• To introduce Sub-GOFA, a novel tool for Sub-Gene Ontology function analysis in clonal mosaicism.
• To leverage semantic similarity for more granular analysis of gene ontology (GO) network structures.
• To identify disease-associated genetic functions by extracting significant differences in sub-GO root-terms.

Main Methods:

• Sub-GOFA analyzes semantic similarity among patients using segmented sub-GO network structures of varying sizes.
• Clustering analysis is performed on these sub-GO structures.
• Significant differences in sub-GO root-terms are extracted to pinpoint disease-associated functions.

Main Results:

• Sub-GOFA effectively measures semantic similarity within clonal mosaicism data.
• The tool successfully extracts disease-associated genetic functions by identifying significant sub-GO root-term differences.
• Validation demonstrated that Sub-GOFA selects a high ratio of cancer-associated genes at an acceptable threshold.

Conclusions:

• Sub-GOFA offers a powerful approach to analyze clonal mosaicism by focusing on sub-gene ontology functions.
• The tool enhances the ability to detect low-frequency genes and elucidate qualitative differences between cancer patients and non-patients.
• Sub-GOFA shows promise for improving cancer gene discovery and patient stratification through advanced bioinformatics analysis.