These are chat archives for dereneaton/ipyrad

Feb 2018
Feb 06 2018 20:04
Hi Deren and Isaac. I have ddRAD, 2x150 data (96 samples, ~220 GB) with which I am trying to identify loci for a RAD capture study. I have been having trouble getting through step 3, running on an HPC (up to 28 cores). The clustering appears to go fairly quickly but slows down dramatically once it reaches >95% to the point that it almost seems to take exponentially longer for each additional percent. The problem may be low quality scores towards the ends of the reads leading to variable length reads (basing this on fastqc results and losing a lot of reads for falling below the minlength). When I increase the quality control parameters to only allow better reads to get through, all steps finish quickly but I get very low numbers of loci per individual. Do you have any suggestions for how to proceed to be able to move past step 3?
Feb 06 2018 21:47
Hello @isaacovercast , I have rad library and did pair end sequencing which gives me R1 and R2 fastq data for rad. I am wondering if only R1 should be provided for rad type? ipyrad demultiplexed the R2 files and combine the few reads with the data demultiplexed from R1. Is there a way to demultiplex pair-end data for rad? Did I miss something? Thanks very much!
Feb 06 2018 23:00
@isaacovercast Hi, Isaac. I have been running several jobs in ipyrad 0.7.17 via qsub on our department's cluster. These jobs were in the middle of running, and running fine, when our system was rebooted. After the system was rebooted, I resubmitted the jobs that had been running, and now they are being terminated with a signal 7 bus error. I contacted our department IT about this and they said that during the system reboot, nodes were upgraded to run kernel 2.6.32-696.18.7.el6.x86_64 (on CentOS 6.9), which is supposed to have some fixes for the recently announced Spectre/Meltdown vulnerabilities. These fixes change what processes see in their address space, which could correlate with these signal 7 bus errors I am seeing. Our IT is wondering whether the new kernel is a problem for ipyrad, and wanted me to find out from the developers what kernel you are using. Can you tell me, what kernel you are using? Thanks!