Global inference of disease-causing single nucleotide variants from exome sequencing data

Area of Science:

• Genetics
• Computational Biology
• Bioinformatics

Background:

• Whole exome sequencing (WES) is crucial for identifying genetic variants in human diseases.
• Current analysis methods struggle with the large number of rare variants and often ignore noncoding regions.
• Existing filtration techniques remove common variants but leave many rare, potentially deleterious variants for further investigation.

Purpose of the Study:

• To develop a novel computational method, Glints, for identifying disease-causing single nucleotide variants (SNVs).
• To enable variant detection in both coding and flanking noncoding regions from exome sequencing data.
• To improve the efficiency and accuracy of genetic variant analysis in human genetics research.

Main Methods:

• Glints integrates 14 functional score types (coding and noncoding predictions) and 9 association scores for disease-relevant genes.
• The method operates on the principle that disease-causing variants affect both the variant and gene levels.
• Large-scale simulation studies using 1000 Genomes Project data were conducted to validate the method.

Main Results:

• Glints demonstrated effectiveness in identifying disease-causing SNVs in both coding and flanking noncoding regions.
• The method showed successful application in two real exome sequencing case studies.
• Simulation studies confirmed the method's capability across diverse genomic regions.

Conclusions:

• Glints is an effective tool for uncovering disease-causing SNVs in both coding and noncoding exonic regions.
• The method's efficacy is supported by simulation and real-world case studies.
• Glints is anticipated to be a valuable asset for human genetics research utilizing exome sequencing data.