Error in step2: Error in type.convert.default(data[[i]], as.is = as.is[i], dec = dec, : long vectors not supported yet: ../../../include/Rinlinedfuns.h:537

[suiyangsun]

Dear Rounak,

I am currently using the FastSparseGRM pipeline and ran into an issue during the step2. My sample size is 450K and pruned bfile has 97,662 variants.

The Error I got:

getDivergence(file.seg,prefix.in,num_threads,divThresh,degree,nRandomSNPs,prefix.out)
[1] "Calculating ancestry divergence for the samples in the seg file"
Error in type.convert.default(data[[i]], as.is = as.is[i], dec = dec, :
long vectors not supported yet: ../../../include/Rinlinedfuns.h:537
Calls: getDivergence -> read.table -> type.convert -> type.convert.default
Execution halted

Here's my code:
####step2:Get ancestry divergence estimates
R CMD BATCH --vanilla '--args --prefix.in /mnt/project/ysui/project/Obe_WES/STAAR/GRM/ukb23156_autosomes_genotype_variant_QCed_extraQC_forGRM --file.seg /mnt/project/ysui/project/Obe_WES/STAAR/GRM/subset/combine.all.seg --num_threads 110 --degree 3 --nRandomSNPs 10000 --prefix.out output.divergence' getDivergence_wrapper.R getDivergence.Rout

Due to my large sample size, I split entire cohort into subset and run king by --projection N in step1 to save time and resources, and then I tried to use this subset seg file for step2 to test, it run successfully! So, do you think it also possible if I use the subset seg file for step2 and then combine all outputs together?

Many thanks,
Yang

[rounakdey]

Could you please share how the first few lines of /mnt/project/ysui/project/Obe_WES/STAAR/GRM/subset/combine.all.seg look like? You can mask any sensitive information. It seems that there is some issues in reading the .seg file. How did you combine the .seg file? One guess is that if you've just combined multiple .seg files using awk or some similar method and the header line is not removed for the second file onward.