Dear Class,
Tom asked me to relay a few updates about homework 6 to save you some frustration.
First, for Part B #1, you may notice that traits with a small sample size have a negative heritability estimate on the Neale Lab website. You should interpret this value as essentially zero.
Second, for Part C #'s 3 and 4, you should first change the beta/OR values so they all indicate increased risk. For beta, this means taking the absolute value (sign should be positive). For OR, this means taking the inverse of any OR that is less than 1 (1/OR). Then you can take the distribution of the effect sizes when these are all in the same direction. This is a standard step for creating polygenic risk scores.
Third, you only need to present your SNPclip results for two chromosomes. You'll notice that SNPclip requires that you enter a list of SNP's that are all on the same chromosome. This means you should sort your GWAS results by chromosome, enter the rsid's for the first chromosome, run SNPclip, then repeat for the SNP's on the next chromosome.