My group has generated Solid 4 targeted capture sequence with initial alignment using Solid's in-house Bioscope tool. We are trying to use GATK for variant calling (for consensus comparisons) but keep running into issues. After troubleshooting several issues, it seems that the fundamental problem is difference in sorting and labeling of chromosomes between Bioscope's BAMs and GATK's reference (although there may be other issues upstream of this one too).
The GATK developer's response to an earlier query was to realign the reads using GATK's reference and thus avoid all downstream problems; however, preliminary efforts twd this don't look promising as Bioscope doesn't like the GATK reference. This series of difficulties has raised the question, "Surely, someone else has performed GATK variant calling on Solid sequenced and aligned data?" So, we are curious if any one could point us a workflow for this process?
In the event that you do not know of a comprehensive workflow for the process, any ad hoc advice would also be helpful. Right now we considering two kludgey solutions to workaround. First, deleting all reads in our BAMs mapping to chrM, chrX and chrY, as our capture library included no sequence from these chr's and they are the ones ordered differently btwn our BAMs and the GATK reference. Second, reordering the bams using the Picard tool reordersam. Any thoughts on the viability of either of these approaches? Are you seeing anything we are missing?