Struggling to use the GATK HaplotypeCaller on my bam files.

Question

[lovegenetics]

Good morning everyone,

I have downloaded the current version on my linux server from https://software.broadinstitute.org/gatk/download/ using the following command:

$ wget "https://github.com/broadinstitute/gatk/releases/download/4.1.2.0/gatk-4.1.2.0.zip"

The software was download and the jar files were made available. The next thing I wanted to do was use the software on my bam files in order to identify the different haplotypes present. I have two samples, and each sample has 5 bam files. I already have other software packages installed to my server like SAMtools, Picard, IGV etc.

However when I ran this command:

$ java -jar gatk-package-4.1.2.0-local.jar -T HaplotypeCaller -R chr86800001-6851705.fasta -I BC01PCR1.sorted.bam -o BC01PCR1_raw_variants.vcf

I get the following error message:

A USER ERROR has occurred: '-T is not a valid command.


***********************************************************************
Set the system property GATK_STACKTRACE_ON_USER_EXCEPTION (--java-options '-DGATK_STACKTRACE_ON_USER_EXCEPTION=true') to print the stack trace.

Next, I tried this command where I just added the genotype mode option and changed the order of commands:

$ java -jar gatk-package-4.1.2.0-local.jar -R chr86800001-6851705.fasta -T HaplotypeCaller -I BC01PCR1.sorted.bam --genotyping_mode DISCOVERY -o BC01PCR1_rawvariants.vcf

And this is the error message I got:

***********************************************************************

A USER ERROR has occurred: '-R' is not a valid command.


***********************************************************************
Set the system property GATK_STACKTRACE_ON_USER_EXCEPTION (--java-options '-DGATK_STACKTRACE_ON_USER_EXCEPTION=true') to print the stack trace.

Can someone please tell me what is the problem what I should have done differently or what do I need to download in order to use GATK. I'm really interested in using the HaplotypeCaller option on my files.

[WouterDeCoster]

Can you see if just gatk is on your $PATH? I believe they now use a wrapper script, and you should not explicitly do java -jar gatk-package-4.1.2.0-local.jar etc etc. If you are following guidelines, make sure they are for the version you are currently using and not for an older version as things change.

[2nelly]

hi alhamidi.reem,

Of course you get this error. If you read carefully the documentation you will notice that the correct way to run any tool of gatk4 (let s say HaplotypeCaller) is:

java -jar gatk.jar HaplotypeCaller

-T parameter is deprecated now. It was used in older versions.

In the second case that you run, you put again -T (which is wrong) and you inverted the correct order of parameters. As I can see from your code, you called gatk and then added the refererence. So in that case gatk was trying to find a tool named -R.

Hope this will help you to run it correctly.

Cheers

[lovegenetics]

Thank you all for your responses. I appreciate it.

Answer 1

[Nicolas Rosewick]

In the GATK folder you should have an executable file named gatk

you should add it to your PATH or create a variable in your script storing the path to this file :

gatk="path/to/the/file/gatk"

genome="path/to/genome/file/" # fasta file used to align the data e.g. hg38 for human
interval="path/to/interval/bed/file"  # required for targeted data e.g. Whole-Exome Sequencing

$gatk HaplotypeCaller \
    --java-options "-Xmx40g -Xms40g" \
    -R $genome \
    -L $interval \
    -O out.vcf \
    -I align.bam \
    -ERC GVCF \
    --max-alternate-alleles 3