Hi members,
I am trying to use the Scalpel indel caller (http://scalpel.sourceforge.net/) to find indels in WGS files from TCGA. But am having trouble with Scalpel running very slowly on original BAM files from TCGA.
I am supplying Scalpel (./scalpel-discovery --single) with a BED file containing a number of 150bp regions within which to search for indels, together with the WGS aligned BAM file and reference genome fasta file. I have left all parameters set to the default. However, I have found that Scalpel runs incredibly slowly on TCGA BAM files that I downloaded directly from Genomic Data Commons (Legacy Archive). It takes ~3 hours to find indels in only 30 150bp regions. However, when I realign the original BAM file (using either Bowtie2 or BWA, with default parameters), Scalpel runs quickly and accurately, taking < 10 minutes to find indels in 80,000 150bp regions. I have seen this problem with files from COAD and LUAD cancer datasets.
Does anyone have any idea why Scalpel might run slowly on TCGA original BAM files only? Or alternatively, any suggestions for things that I could try to troubleshoot this issue?
I have attached a file with analysis metadata for a sample TCGA BAM file as this gives the alignment pipeline that was used on the original files.
Thank you in advance for any help!
I am trying to use the Scalpel indel caller (http://scalpel.sourceforge.net/) to find indels in WGS files from TCGA. But am having trouble with Scalpel running very slowly on original BAM files from TCGA.
I am supplying Scalpel (./scalpel-discovery --single) with a BED file containing a number of 150bp regions within which to search for indels, together with the WGS aligned BAM file and reference genome fasta file. I have left all parameters set to the default. However, I have found that Scalpel runs incredibly slowly on TCGA BAM files that I downloaded directly from Genomic Data Commons (Legacy Archive). It takes ~3 hours to find indels in only 30 150bp regions. However, when I realign the original BAM file (using either Bowtie2 or BWA, with default parameters), Scalpel runs quickly and accurately, taking < 10 minutes to find indels in 80,000 150bp regions. I have seen this problem with files from COAD and LUAD cancer datasets.
Does anyone have any idea why Scalpel might run slowly on TCGA original BAM files only? Or alternatively, any suggestions for things that I could try to troubleshoot this issue?
I have attached a file with analysis metadata for a sample TCGA BAM file as this gives the alignment pipeline that was used on the original files.
Thank you in advance for any help!