Hello Everyone, I'm new to GWAS so forgive me for asking naive question. I have my summary statistics(From Linear mixed model) looks like below:
CHROM GENPOS ID ALLELE0 ALLELE1 A1FREQ INFO N TEST BETA SE CHISQ LOG10P
Allele0 is the reference allele and allele1 is the alternative allele. I wonder how to calculate R^2 from this summary. The GWAS model used normal approximation, 𝜒^2 chi-square to estimate the P-value.
I saw different equations such as SNP (PVE)=2p(1-p)*beta^2
, and SNP (PVE) = [2*(beta^2)*MAF*(1-MAF)]/[2*(beta^2)*MAF(1-MAF)+((se(beta))^2)*2*N*MAF*(1-MAF)]
What's the best software to analyze the percentage of phenotypic variation explained by the identified SNPs?
Thank you very much!
Given the p-value and direction of effect, you can use the equation stated here: https://en.wikipedia.org/wiki/Pearson_correlation_coefficient#Testing_using_Student's_t-distribution to convert the t-statistic to the r2. That is what's used by lassosum.
Thank you so much Sam! You are always a great helper. I realized that my GWAS model uses normal approximation 𝜒^2 chi-square to estimate the P-value. I'm getting confused about the relations now.
Log in to answer this question.
I am not sure about where the second expression comes from.
You can read the response I gave in the following related post, which motivates the first expression. Note however that your trait has to have unit variance.
How to estimate the percent phenotypic variation explained by the significant SNP from a GWAS study?