How To Measure The R2 Value Of Impute2 Results Using The True Genotypes (Simulated Data) As Comparison

Question

[tomislav.ilicic]

Hello,

I am new to imputation and I am wondering if there are any scripts already avaialble to measure r2 value of impute2 results using the true genotypes (simulated data) as comparison.

When I get output from impute2 like that: --- rs_id_0 42 A T 0.975 0.025 0

I wonder what the genotypes are. Is it in that order: A/A A/B B/B ? Many thanks, Tomi

Answer 1

[zx8754]

Read on - "Posterior genotype probabilities", here is a good intro: http://www.uni-kiel.de/medinfo/lehre/vorlesungen/genepi/folien/genepi_20.pdf

For concordance tests, see: http://mathgen.stats.ox.ac.uk/impute/impute_v2.html#concordance_tables

Check GTOOL for merging gen files and if needed converting gen files to hard calls at some threshold: http://www.well.ox.ac.uk/~cfreeman/software/gwas/gtool.html