seGMM: A New Tool for Gender Determination From Massively Parallel Sequencing Data.

Area of Science:

• Genetics
• Bioinformatics
• Molecular Diagnostics

Background:

• Gender concordance is vital in clinical genetic testing to prevent diagnostic errors.
• Existing gender inference tools are not optimized for targeted gene sequencing (TGS) data.
• Sex chromosomal abnormalities can impact molecular diagnosis and treatment.

Purpose of the Study:

• To validate the seGMM tool for accurate gender inference from TGS data.
• To assess seGMM's capability in identifying sex chromosomal abnormalities.
• To compare seGMM's performance against existing gender inference tools.

Main Methods:

• Utilized unsupervised clustering (Gaussian mixture model) for gender determination.
• Aligned sequencing reads to identify sex chromosomal abnormalities.
• Validated seGMM on diverse TGS, whole exome sequencing (WES), and whole genome sequencing (WGS) datasets.

Main Results:

• seGMM achieved >99% gender-inference accuracy across multiple datasets.
• Demonstrated superior performance compared to PLINK, seXY, and XYalign for TGS data.
• Successfully identified sex chromosomal abnormalities, confirmed by amelogenin analysis.

Conclusions:

• seGMM is a highly accurate tool for gender determination and sex chromosomal karyotyping from sequencing data.
• seGMM enhances quality control in clinical genetic testing, especially for TGS.
• The tool shows significant potential for improving molecular diagnosis and treatment decisions.