Calling indels (GATK)

[gustavo-miranda]

Hi Brant,

I am using the seqcap_pop scripts within some of the phyluce tools and I am having a problem with indel calling.

When calling indels, we should provide the path to match-contigs-to-probes fasta file (-R) and to the merged bam file (-I). In my match-contigs-to-probes folder I have several fasta files, one for each individual of SpeciesA. In my merge-bams folder, I have only one single bam file (and its index) with bams of all specimens merged into the same file.

The problem I am having is that when I set the path/to/match-contigs-to-probes/SpeciesA_specimen050.fasta it only reads information of SpeciesA_specimen050 in the bam file and gives me an error for all other specimens. Here is how the output looks like:

MESSAGE: Badly formed genome location: Contig uce-4998_montanus051 given as location, but this contig isn't present in the Fasta sequence dictionary

Is there a way to use a command that makes GATK reads a file with a list of paths to all fastas making the the program run all specimens at once?

Here are the commands I am using: java -Xmx2g -jar ~/anaconda/GenomeAnalysisTK-3.3-0/GenomeAnalysisTK.jar \ -T RealignerTargetCreator \ -R /path/to/4_match-contigs-to-probes/Genus_species.fasta \ -I /path/to/7_merge-bams/Genus_species.bam \ --minReadsAtLocus 7 \ -o /path/to/8_GATK/Genus_species.intervals

This might be of interest of @mgharvey too.

Thank you. Gustavo

[brantfaircloth]

Hi Gustavo,

The way the GATK code is written, at present, requires that you only have a single reference assembly to which you have aligned your individual data. So, you cannot have the code align data for multiple individuals to multiple references.

Hope that helps clarify, -b

[gustavo-miranda]

Hi Brant,

Yes, it does clarify. Thank you!

Gustavo