The frontline support team will be unavailable to answer questions on April 15th and 17th 2019. We will be back soon after. Thank you for your patience and we apologize for any inconvenience!

GATK and 454 data

MatteodiggMatteodigg ItalyMember
edited August 2015 in Ask the GATK team


I'm new in this forum and variant calling. I have read most of your posts and guides but I didn't find a satisfactory answer to my doubts. I'm working on a pipeline for germline variants identification in human targeted sequencing. I have to develop a good pipeline for Illumina and 454 platforms. I used GATK for my Illumina datas, following the best practice workflow according to my datas and it seems to be ok. However, I have some doubts regarding to 454 analysis. I'd like to have some clarification about it.
First of all, I'm working on small target sequencing (more or less 120k bases per sample) on 6 samples. Working on 454 datas, I cannot use IndelRealigner because of Indel problem on 454 technology. In the same way, I cannot use BQSR as well as VQSR because I have a small amount of datas (this problem occurs on Illumina platform too). For this reason, you suggest using hard filtering with the VariantFiltration tool according to my data and my specifications. In few words, GATK "works fine" on WGS and WES (especially with multi-sampling) whereas, generally, I have a good coverage and quality score. Conversely, if I work with targeted sequencing, especially with 454 technology (due to poor mean coverage and quality score), results could be less clear and open to interpretation according to my data. I hope my doubts are clear. So, in coclusion, I'd like to use GATK for both 454 and Illumina and I need this evaluation and explanation to clarify if I can use GATK on both "poor" and "good" data. I hope I was clear and this is not a too stupid question.
Thank you in advice for your help!

Best Answers


Sign In or Register to comment.