search

zhengxwen / SNPRelate

R package: parallel computing toolset for relatedness and principal component analysis of SNP data (Development version only)
http://www.bioconductor.org/packages/SNPRelate
101 stars 25 forks source link

details of snpgdsLDpruning required #48

Closed thierrygosselin closed 6 years ago

thierrygosselin commented 6 years ago

Hi,

In the function documentation of SNPRelate::snpgdsLDpruning, details are given on the sliding window implementation, but what is lacking in the detail section: the pruning process.

e.g. for a pair of SNPs in linkage lets say the r = 0.9. The pruning process will keep only 1 SNP.

Question how did you implement this selection process ? Is it random selection or is it implemented based on SNP with higher MAF or lower missing rate? (because you do start the detail section with those statistics).

Best Thierry

zhengxwen commented 6 years ago

It is a random selection.

thierrygosselin commented 6 years ago

thanks for the details