chromosomewise lassosum and pseudovalidate error

[ghost]

Dear tshmak

Hello, I am running the lassosum.pipeline in unix but I have some errors related to this package.

First chromosome-wise running is not working in some chromosomes.

I am using the code as follows.

out = lassosum.pipeline(cor=cor,chr=ss.df$CHR,pos=ss.df$BP,A1=ss.df$ALT,A2=ss.df$REF,ref.bfile="/data/pkg36/lassosum/data/refpanel",test.bfile=test.bfile,LDblocks="ASN.hg19")

out22 = lassosum.pipeline(cor=cor,chr=ss.df$CHR,pos=ss.df$BP,A1=ss.df$ALT,A2=ss.df$REF,ref.bfile="/data/pkg36/lassosum/data/refpanel",test.bfile=paste0(test.bfile,"_chr22"),LDblocks="ASN.hg19") out21 = lassosum.pipeline(cor=cor,chr=ss.df$CHR,pos=ss.df$BP,A1=ss.df$ALT,A2=ss.df$REF,ref.bfile="/data/pkg36/lassosum/data/refpanel",test.bfile=paste0(test.bfile,"_chr21"),LDblocks="ASN.hg19") out20 = lassosum.pipeline(cor=cor,chr=ss.df$CHR,pos=ss.df$BP,A1=ss.df$ALT,A2=ss.df$REF,ref.bfile="/data/pkg36/lassosum/data/refpanel",test.bfile=paste0(test.bfile,"_chr20"),LDblocks="ASN.hg19")

out22 and out21 works fine, but out20 gets an error like this:

But lassosum.pipeline for whole chromosomes; out works.

Second, as I merged the outcomes, in my case out22 and out21, validate works but pseudovalidate doesn't work.

out = merge(out22,out21) v = validate(out) v = pseudovalidate(out)

I tried to figure out the reason for these errors, but I am lost.

If you have any comments or tips for fixing these, please let me know.

Best Wishes

[tshmak]

How many SNPs are there in each of the chromosomes?

Have you made sure that these also exist in test.bfile?

It sounds as if there are only a few SNPs.

[jyjo43043]

I summarized the number of SNPs in each of the chromosomes in summary statistics file and test.bfile below.

I matched the SNPs in these two files to have the same number of SNPs and same SNPs.

It seems there exist enough SNPs in common, I think.

[tshmak]

The error message suggests that there are no SNPs left after matching the summary stats to test.bfile. It'd be difficult for me to figure out why without having your data. Would you be able to debug to see why that's the case? See e.g. https://support.rstudio.com/hc/en-us/articles/205612627-Debugging-with-the-RStudio-IDE