search

ChongWuLab / SUMMIT

Summary-level Unified Method for Modeling Integrated Transcriptome
10 stars 1 forks source link

step2_pos.R - problem with the Z score of TWAS #1

Open MinghanQ opened 9 months ago

MinghanQ commented 9 months ago

Hi, I am currently attempting to apply the SUMMIT method for TWAS analysis. While studying the software source code, I noticed that in 'step2_pos.R’, on line 271, the p-value is calculated by dividing z.twas by the square root of r2.twas. According to the formula in your original article, it seems that this value is the Z score for TWAS. However, on line 272, z.twas is directly output as the Z score. Why is this the case? Could it be a misunderstanding on my part?"

melothemightyone commented 9 months ago

Hi MinghanQ,

Your observation is keen and correct. The object z.twas / sqrt(r2.twas) is the Z-score that should be outputted. We will fix this issue in an updated version of SUMMIT in the near future.