Detecting genomic regions associated with a disease using variability functions and Adjusted Rand Index

Area of Science:

• Comparative genomics
• Bioinformatics
• Computational biology

Background:

• Identifying functional genomic regions in multiple sequence alignments is a key challenge in comparative genomics.
• Existing methods often overlook the link between functional regions and external species-specific data like carcinogenicity.
• Prior work introduced methods incorporating external evidence for disease-related genomic region identification.

Purpose of the Study:

• To develop a novel algorithm for detecting genomic regions associated with specific diseases.
• To introduce and evaluate new variability functions and a bipartition optimization procedure.
• To assess the algorithm's performance using metrics like the Adjusted Rand Index (ARI) against known species classifications.

Main Methods:

• Development of two novel variability functions and a bipartition optimization procedure.
• Validation using the Adjusted Rand Index (ARI) to correlate detected regions with species classifications (e.g., carcinogenicity, invasivity).
• Assessment of predictive power on both synthetic and real biological data.

Main Results:

• The new algorithm successfully identifies genomic regions potentially associated with diseases.
• Results indicate that no single detection function is optimal for all evolutionary scenarios; at least three may be necessary.
• The proposed functions, even without prior knowledge, achieve results comparable to existing methods that utilize such information.

Conclusions:

• The developed algorithm offers a robust approach to disease-associated genomic region detection.
• The study highlights the utility of multiple variability functions for comprehensive analysis.
• Applied to specific examples, the algorithm confirmed known features of Neisseria meningitidis and Human Papilloma Virus (HPV).