IMPUTE2 -merge_ref_panels

Question

[raalsuwaidi]

Hi all,

i am trying to use IMPUTE2 with 2 reference panels to be merged. i am applying the code as per the example provided on IMPUTE2 page. but somehow the merged reference panel doesnt get produced (the REF file as per the example). any ideas?

for simplicity i am putting the code

impute2 -merge_ref_panels  \
-m  genetic_map_chr1_b36.txt \
-h ALL.chr1.phase3_shapeit2_mvncall_integrated_v5a.20130502.genotypes.vcf.gz.hap.gz \  local_chr1.vcf.hap.gz \ 
-l  ALL.chr1.phase3_shapeit2_mvncall_integrated_v5a.20130502.genotypes.vcf.gz.legend.gz \
 local_chr1.vcf.legend.gz \
-g fileToBeImputed.QC.CHR1.vcf.gen.gz   \
-int 20.4e6 20.5e6 \
-Ne 20000 \
-merge_ref_panels_output_ref   REF  \
-o chrm1_output

I cant find the REF file!!!! am i doing something wrong?

[Kevin Blighe]

I cant find the REF file!!!!

What do you mean?

[raalsuwaidi]

I mean the output from the below statement

-merge_ref_panels_output_ref REF \

[Kevin Blighe]

You should not have any file called REF. That parameter defines the name prefix of the merged panel output files. Please see: https://mathgen.stats.ox.ac.uk/impute/panel_merging_options.html

Kevin

[raalsuwaidi]

i changed it into -merge_ref_panels_output_ref mergedref and i still cannot see the mergedref file

[Kevin Blighe]

Please show all status, warning, and error messages

[raalsuwaidi]

there is no error or warning, but in the summary file the last entry is as follows:

---

Run log

RESETTING PARAMETERS FOR "SURROGATE FAMILY" MODELING -setting mutation matrices -setting switch rates -setting hole-filling switch rates

diploid sampling success rate: 0.000

haploid sampling success rate: (no haploid sampling performed)

i dont know what that means

[Kevin Blighe]

If there are no warnings or errors, and the program indicates that the command was successful, then there should be output files with prefix mergedref

What output is actually produced?

[raalsuwaidi]

these are the output files produced

imput2_combined_merged_chrm1
imput2_combined_merged_chrm1_info imput2_combined_merged_chrm1_info_by_sample imput2_combined_merged_chrm1_summary imput2_combined_merged_chrm1_warnings

the warning file is empty. and the summary file has an output of concordance value 99.5.

the code used is:

impute2 -merge_ref_panels 
-m  genetic_map_chr1_b36.txt
-h ALL.chr1.phase3_shapeit2_mvncall_integrated_v5a.20130502.genotypes._normalized.vcf.hap.gz 
local_chr1_normalized.vcf.hap.gz 
-l ALL.chr1.phase3_shapeit2_mvncall_integrated_v5a.20130502.genotypes._normalized.vcf.legend.gz 
local_chr1_normalized.vcf.legend.gz 
-g merged.QC.CHR1.vcf.gen.gz 
-int 20.4e6 20.5e6 
-Ne 20000 
-merge_ref_panels_output_ref mergedpanel 
-o imput2_combined_merged_chrm1

the file that should be the output of -merge_ref_panels_output_ref mergedpanel is not in the folder with the rest of the files

i have no idea what is wrong.

Answer 1

[raalsuwaidi]

i am adding the solution as i finally fixed it and maybe it would help some one else.

it seems to generate the merged panel you should not add a target to be imputed. so removing the -g parameter resulted in obtaining the merged panel