Bridging heterogeneous mutation data to enhance disease gene discovery.

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Genome-wide association studies (GWAS) identify gene-disease links but can miss significant mutations due to stringent statistical thresholds.
• Mutation data from literature provides functional context (e.g., gain/loss of function) often missed by GWAS.
• Integrating heterogeneous mutation data is crucial for comprehensive disease gene discovery.

Purpose of the Study:

• To develop a computational pipeline (GDAMDB) for bridging heterogeneous mutation data.
• To recover false-negative GWAS mutations by integrating literature-reported functional evidence.
• To enhance the discovery of novel disease-associated genes.

Main Methods:

• Developed a Gene-Disease Association prediction by Mutation Data Bridging (GDAMDB) pipeline.
• Employed a statistical generative model to learn mutation association and type distributions.
• Integrated GWAS data with mutation information extracted from scientific literature via text mining.

Main Results:

• Applied GDAMDB to Alzheimer's disease (AD), predicting 79 associated genes.
• Of the 79 genes, 12 were identified in the original GWAS, and 60 were supported by other GWAS or literature.
• Identified novel AD-associated genes beyond existing discoveries.

Conclusions:

• Bridging heterogeneous mutation data significantly contributes to novel disease-related gene discovery.
• The GDAMDB pipeline effectively enhances GWAS-based gene association discovery by integrating text mining results.
• This approach offers a powerful strategy for uncovering complex gene-disease relationships.