whats the dataset on the NCBI website, 1000 genomes? you want to test against controls from the same chip, otherwise you have no way of knowing if a statistical difference is due to phenotype or due to chip effects. even from the same chip, you need to run some QC to account for batch effects.Ghost wrote: ↑Sun Feb 07, 2021 4:59 pmI should clarify, I posted the raw data the program generated. Some time in the near furure I'm going to run it against the frequencies posted on the NCBI website.
RE rs1042778: I don't see it in the list of snps snpedia listed on the v4 chip, so I didn't grab the frequencies from the data. I can go back and grab it later if you'd like.
how many snps were on the chip before qc? I take it this is unimputed data?