dindel how to merge library files in pooled setting

[slowsmile]

Dear all
I am new to dindel and I tried to call indels from a pool of bam files. I followed their manual, which ran smoothly until step 3 when I used the --bamFiles option to analyze all bamfiles together.

Then I don't know which library file should I feed in (in first step, it produces a library file for each bam file). Then I merged the library lifes by row and fed into step 3, only to see errors:

Storing library: unknown
Error: libName: unknown
Duplicate library IDs.
Exception: Library error

can any one tell me how to produce merged library files for multiple bam files?

Thanks a lot

[deyler]

duplicate library ID error in Dindel

While your need may be over, I'm going to post the solution as a reference for the community.

When Dindel reads bam files, it looks in the @RG lines for a library name, specfied with the LB: tag according to sam spec. If it's not there, it uses 'unknown' or 'dindel_default.' See Read.hpp in the dindel directory.

You'll find the library name in the header line of the foo.libraries.txt file generated by dindel getCIGARindels.

#LIB     dindel_default

If you run the pooled analysis, it wants all your libraries.txt files to have different library names (because you did pooled sequencing of several different libraries).

To solve this, either
1) edit the *.libraries.txt files to have unique library names in their header lines, or
2) edit your bamfiles to have unique LB: entries in the @RG line. You can do this with the samtools reheader tool.

[sandipkale]

Hello All,

I am running Dindel on 100 samples. I have completed all the steps upto genotype likelihood calling and got the glf file. However, I need results in vcf/hapmap format.
Could you please help me in converting glf to vcf/hapmap format?

Thank you very much in advance

Sandip

[deyler]

perhaps a python script

My recollection is that there are some python scripts that come with Dindel that you can use to convert between file formats. Unfortunately, I can't provide more information because I did not use Dindel in my finished pipeline.

[sandipkale]

Hello Dr. deyler,

Thank you very much for reply. I checked the dindel manual. There are four helper scripts. We can use mergeout*.py script for getting VCF, however, as the glf output I got using makegenotype*.py is in different format, the above scripts can't be used.

I also called Indels separately and would like to merge the vcf files. Is there any way to merge the vcf files other than vcf-merge?

Thank you in advance

With best regards