plinkQC

plinkQC integrates ancestry inference, sample selection, and quality control for population genetics analyses.

Key Features:

• Ancestry Inference: Determines ancestry of study samples via a pre-trained random forest classifier that reaches 98% performance accuracy with just 5% of marker overlap between reference and user data.
• Sample Selection: Obtains the maximal set of unrelated study samples using a graph-based pruning method that incorporates both relationship estimates and sample quality.
• Quality Control: Bundles per-individual and per-marker quality control checks into functions that return both the quality controlled dataset and a quality control report.
• PLINK Integration: Leverages PLINK to streamline common tasks and improve data quality.

Scientific Applications:

• Population Genetics: Provides rigorous controls for sample and marker quality required for population genetic analyses.
• Ancestry Identification: Identifies ancestry of study samples when reference and user data have limited marker overlap.
• Relatedness Analysis: Processes sample relatedness to obtain optimal sets of unrelated individuals for downstream analyses.
• Large-Scale Cohort Processing: Demonstrated on the 1000 Genomes project, retaining additional samples compared to publicly available exclusion lists.

Methodology:

plinkQC uses a pre-trained random forest classifier for ancestry inference and a graph-based pruning method for sample selection that accounts for relationship estimates and sample quality.