Error in if (a1 != a2) .LOG("The effect column was determined to be coded as an odds ratio (OR) for the ", : missing value where TRUE/FALSE needed

8 views
Skip to first unread message

Royce Clifford

unread,
Sep 28, 2025, 1:21:08 PM (3 days ago) Sep 28
to Genomic SEM Users

Good morning:

I am using brain volume GWAS summary statistics from Tissink, et al. for GSEM. When I try to munge, I receive the following error: Error in if (a1 != a2) .LOG("The effect column was determined to be coded as an odds ratio (OR) for the ",  :missing value where TRUE/FALSE needed 

Here are a few lines from the GWAS file:

SNP                    CHR          BP A2    A1           MAF     N  INFO    BETA    SE       P

rs10000000 4      40088896 A     T     0.0601 26484 0.982  520.   451. 0.249

rs10000010 4      21618674 T     C     0.487  26044 0.986   48.3  203. 0.812

rs1000002  3     183635768 C     T     0.478   26643 1      285.     201. 0.155

rs10000023 4      95733906 T     G     0.424   25709 0.981 -296.     207. 0.153

rs1000003  3      98342907 A     G     0.159   26535 0.994   -3.49   281. 0.990

rs10000033 4     139599898 T     C     0.452   26454 0.993    0.795  201. 0.997

 

The BETAS were normalized – is that a problem?

In the resultant .sumstats.gz file, there appeared to be sufficient lines, 1144480, I looked for Inf values and NA values - none noted. Here is some of the sumstats file.

SNP      N           Z            A1         A2                      

rs1000000       26638 1.81     A            G

rs10000010    26044 -1.68     C           T

rs1000002       26643 1.67      T            C

rs10000023    25709 0.933   G           T

rs1000003       26535 1.97      G           A

rs10000033    26454 -0.350 C            T

How can I resolve this?

Thanks.

Reply all
Reply to author
Forward
0 new messages