search

Stahl-Lab-MSSM / coloc2

Upgrade to coloc
13 stars 7 forks source link

Error when running coloc.eqtl.biom #4

Open ibrahiaa opened 1 year ago

ibrahiaa commented 1 year ago

Hi,

After formating the biom.df and eqtl.df objects and adding a $type column to each, I got two data frames looking this, which seems fine:

head(biom.df) SNPID CHR POS PVAL BETA SE A1 A2 N type 1 rs141709959 1 751456 0.690541 0.1593330 0.400583 T C 59225 quant 2 rs28527770 1 751756 0.924024 -0.0214828 0.225478 T C 59225 quant 3 rs201062411 1 752308 0.869061 0.0605275 0.367507 I D 59225 quant 4 rs146277091 1 752478 0.869175 0.0604588 0.367412 G A 59225 quant 5 rs3094315 1 752566 0.923839 0.0183066 0.191672 G A 59225 quant 6 rs149886465 1 752617 0.866254 0.0617738 0.367130 C A 59225 quant

head(eqtl.df) SNPID CHR POS PVAL BETA SE ProbeID A1 1 rs143225517 1 751756 8.375668e-03 0.03676353 0.01388336 ENSG00000228794 T 2 rs143225517 1 751756 7.111666e-03 0.08623658 0.03189537 ENSG00000236239 T 3 rs143225517 1 751756 7.795206e-03 -0.06456827 0.02415937 ENSG00000237491 T 4 rs3094315 1 752566 2.966613e-03 -0.03436617 0.01150450 ENSG00000228794 G 5 rs3094315 1 752566 1.986647e-05 0.08544778 0.01981848 ENSG00000237491 G 6 rs3131972 1 752721 8.914779e-03 -0.02999908 0.01142151 ENSG00000228794 A A2 F N MAF type 1 C 0.8529771 467 0.1470229 quant 2 C 0.8529771 467 0.1470229 quant 3 C 0.8529771 467 0.1470229 quant 4 A 0.1892925 467 0.1892925 quant 5 A 0.1892925 467 0.1892925 quant 6 G 0.1921606 467 0.1921606 quant

However, when running coloc.eqtl.biom using default parameters, I get this error message, not sure what is causing it:

132 SNPs out of 139 had the correct alleles, discarding SNPs without the correct alleles There are 0 duplicated SNP names in the data ENSG00000236239: 132 snps in both biomarker and eQTL data. From: 751756 To: 1874837 Processing dataset Error in d$type == "quant" : comparison (1) is possible only for atomic and list types In addition: Warning message: In coloc.eqtl.biom(eqtl.df = eqtl.df, biom.df = biom.df, p12 = 1e-06, : NAs introduced by coercion

clagiamba commented 1 year ago

Hi, did you check if you have: missing, NA, Inf, something not numeric or=0 for PVAL, BETA, SE ?

ibrahiaa commented 1 year ago

Hi,

Thanks for the quick reply. I checked the columns you mentioned and they are all fine in both datasets, no missing, NA or Inf values. From the error message it seems something with the 'type' column is not right?

ibrahiaa

clagiamba commented 1 year ago

And no pvalue == 1? The type="quant" should be fine, just check that it is character?

ibrahiaa commented 1 year ago

Type is of class character. PVAL column in biom.df contained a few 1 values. I replaced all these with 0.999, but got the same error.

ibrahiaa commented 1 year ago

I found that the biom.df file had som "CHR" numerical values other than 1:22. By removing these, the warning message disappears. But the 'Error in d$type == "quant"' remains.