GemTools

GemTools models genetic ancestry from SNP data using spectral graph theory and projection techniques to decompose large genotypic datasets and analyze population structure.

Key Features:

• Efficient computation: Uses spectral graph theory and a divide-and-conquer decomposition strategy to break large SNP/genotypic datasets into smaller components for faster processing.
• Accurate population clustering: Identifies major genetic clusters and subclusters without prior knowledge of individual origins and can discern within-population variation (93–95% of genetic diversity in a study of 1056 individuals from 52 populations using 377 autosomal microsatellite loci).
• Epidemiological correlation: Correlates genetic clusters with self-reported ancestry to support assessment of epidemiological risks and to inform stratification in association studies.

Scientific Applications:

• Population structure analysis: Analysis of human population structure from SNP and genotypic data using spectral decomposition and projection methods.
• Genetic diversity assessment: Quantification and characterization of within- and between-population genetic variation, exemplified by analyses of 377 autosomal microsatellite loci across 52 populations.
• Epidemiology and association studies: Integration of inferred genetic clusters with self-reported ancestry to assess epidemiological risk factors and improve case-control matching in association analyses.

Methodology:

Spectral graph theory, projection techniques, and a divide-and-conquer decomposition strategy applied to genotypic data (SNPs and autosomal microsatellite loci) for population-structure analysis.