zhanxw
commented
7 years ago Hi Zishan,
The error message reported line numbers (8869314). Can you please check line 8869314 and compare that with previous line and the next line? I would try to generate a small VCF file with line 8869310 to 8869320, re-run rvtests and see if the problem disappears. If not, the problem is probably the input file. You can attach here, and I will check it.
If you don't perform group-based tests, your command will be fine.
zishang30
dajiangliu
Dear Xiaowei
I have checked single chrosome vcf.gz file and bgziped chr.all vcf.gz file for generating kinship matrix. It seems that in chr.all vcf.gz file some VCF line does not have correct column number as follows:
04-30 14:54 Line [ 8869314 ] does not have correct column number, exiting! 04-30 14:54 Current line has 2251 columns. 05-02 23:51 Line [ 16287134 ] does not have correct column number, exiting! 05-02 23:51 Current line has 861 columns.
Because we already checked single VCF.gz, so this error may be due to the bgzip code for bgziping all chromosome ?
The code for bgzip all chromosome are as follows: (zcat chr1_notmonomorph.vcf.gz; zgrep -v '^#' chr2_notmonomorph.vcf.gz; zgrep -v '^#' chr3_notmonomorph.vcf.gz; zgrep -v '^#' chr4_notmonomorph.vcf.gz; zgrep -v '^#' chr5_notmonomorph.vcf.gz; zgrep -v '^#' chr6_notmonomorph.vcf.gz; zgrep -v '^#' chr7_notmonomorph.vcf.gz; zgrep -v '^#' chr8_notmonomorph.vcf.gz; zgrep -v '^#' chr9_notmonomorph.vcf.gz; zgrep -v '^#' chr10_notmonomorph.vcf.gz; zgrep -v '^#' chr11_notmonomorph.vcf.gz; zgrep -v '^#' chr12_notmonomorph.vcf.gz; zgrep -v '^#' chr13_notmonomorph.vcf.gz; zgrep -v '^#' chr14_notmonomorph.vcf.gz; zgrep -v '^#' chr15_notmonomorph.vcf.gz; zgrep -v '^#' chr16_notmonomorph.vcf.gz; zgrep -v '^#' chr17_notmonomorph.vcf.gz; zgrep -v '^#' chr18_notmonomorph.vcf.gz; zgrep -v '^#' chr19_notmonomorph.vcf.gz; zgrep -v '^#' chr20_notmonomorph.vcf.gz; zgrep -v '^#' chr21_notmonomorph.vcf.gz; zgrep -v '^#' chr22_notmonomorph.vcf.gz; zgrep -v '^#' chrX_notmonomorph.vcf.gz)| bgzip -c > HRC_chrall.vcf.gz
Is this correct?
In addition, another thing came to my mind: our statistic plan ask us to use follow command to bgzip and index VCF file: (grep ^"#" $your_old_vcf; grep -v ^"#" $your_old_vcf | sed 's:^chr::ig' | sort -k1,1n -k2,2n) | bgzip -c > $your_vcf_file tabix -f -p vcf $your_vcf_file
this command are used for group-based rare variant tests.
If we don't conduct group-based rare variant tests, should we just use command as follow: bgzip -c file.vcf > file.vcf.gz tabix -p vcf file.vcf.gz
Thank you very much!
Best regards
Zishan