Note: please only do this if you have been explicitly asked to do so.
Scenario:
You posted a question about a problem you had with GATK tools, we answered that we think it's a bug, and we asked you to submit a detailed bug report.
Here's what you need to provide:
- The exact command line that you used when you had the problem (in a text file).
- The full log output (program output in the console) from the start of the run to the end or error message (in a text file).
- A snippet of the BAM or VCF file if applicable and the index file associated with it.
- If you used a non-standard reference (i.e. not available in our resource bundle), we need the .fasta, .fai, and .dict files for the reference.
- Any other relevant files such as recalibration plots
A snippet file is a slice of the original BAM or VCF file that contains the problematic region and is sufficient to reproduce the error. We need it in order to reproduce the problem on our end, which is the first necessary step to finding and fixing the bug. We ask you to provide this as a snippet rather than the full file so that you don't have to upload (and we don't have to process) huge giga-scale files.
Here's how you create a snippet file:
- Look at the error message and see if it cites a specific position where the error occurred.
- If not, identify what region caused the problem by running with
-L
argument and progressively narrowing down the interval. - Once you have the region, use PrintReads with
-L
to write the problematic region (with 500 bp padding on either side) to a new file -- this is your snippet file. - Test your command line on this snippet file to make sure you can still reproduce the error on it.
If the error is happening when you're running on a VCF instead of a BAM, do the same thing but with SelectVariants instead of PrintReads.
And finally, here's how you send us the files:
- Put all those files into a
.zip
or.tar.gz
archive. - Upload them onto our FTP server with the following credentials:
location: ftp.broadinstitute.org username: gsapubftp password: 5WvQWSfi
- Post in the original discussion thread that you have done this.
- Be sure to tell us the name of your archive file!
1 comment
when I use GATK tool HaplotypeCaller,I find a bug: When the two mutations are very close,The results are very inaccurate。It is mainly reflected in two aspects:
(1)This may cause one of the mutations to be missed
For this two mutation,only one was detected by GATK.
(2)One of the very close mutations in the results is a false positive
For example:These two mutations,mutations 1 is a false positive,mutations 2 is True positive,but the qual of this mutations 2 is also very low,just like mutations 1.This mutations 2 may result in being filtered。
How can gatk avoid this situation?Thank you very much!
Please sign in to leave a comment.