LAPSTRUCT: a geometric approach to describe population structure

LAPSTRUCT is a free program to describe population structure using biomarker data ( typically SNPs, CNVs etc.) available in a population sample. The main features different from PCA are: (1)geometrically motivated and graphic model based; (2)robustness of outliers.

Current release: version 1.0 (released in December, 2009)

New Feature: Sparse Laplacian eigenfunctions (coming)

Documentation: [PDF]

Format of input genotype file: [Example] HGDP panel consisting of 800 most informative SNPs of 940 unrelated samples

Download: the program in R, click here

Contact: junzhang at galton.uchicago.edu

References:
1. Zhang J, Niyogi P and McPeek MS (2009) "Laplacian eigenfunctions learn population structure". PLoS ONE, 4(12): e7928. doi:10.1371/journal.pone.0007928 [link]

2. Zhang J, Weng C and Niyogi P (2009) "Graphical analysis of population structure on rheumatoid arthritis data". BMC Proceedings, 3(Suppl 7):S110. [link]

3. Zhang J (2009) "Ancestral informative marker selection and population structure visualization using sparse Laplacian eigenfunctions", under review. [link]
4. Omri j (2012) "The cumulativeeffectofgeneticmarkersonclassificationperformance", J of Theoretical Biology. 293,206-218 [link]