My genome assembly was running fine with 22% completed on the 'Assembly' step. It had already taken 72 hours to get here. My system had to be shutdown for some critical security updates, and I had to interrupt the Minia run. Anyway to re-start from where I aborted, or back to square one?
•
written
by
avismirabilis •
Hi, due to the use of HDF5, the intermediate files do not get properly saved during a run, hence it's generally not possible to restart a minia run. What sort of organism took 72+ hours to assemble? This might be a good occasion to use the latest version of minia (https://github.com/GATB/minia, compile from sources) for improved quality of results.
The genome is a of a Crane species - approx 1.3Gb expected genome size. We have about 800 million reads, so I was wondering if that was the reason for how long it was taking.
Are you running it on a network drive? also, how many threads and free disk space?
Nothing unusual regarding setup. linux server (not cluster), 80 cores with 60 TB free space, and 512Gb RAM.
I took you advice and ran the latest version of Minia- actually using the GATB pipeline. That one completed through all K's within 48 hours. Strange phenomenon was that i got better assembly stats with a single pair of reads (~100 million reads), but was really poor with additional reads and mate pairs added. Perhaps that's a discussion for another thread.
Good to know the assembly completed with the latest version. Did you compare contigs between those two runs, or scaffolds?
Log in to answer this question.
