I have already used Pilon to improve a genome assembly (400Mbp) using .bam alignments of a RADseq library. When I ran the program it took around 12h to finished the entire correction. However, I'm using Pilon again to improve another (shorter 273 Mbp) DeNovo genome and the program has been running for more than 57h. Is it normal? I'm not sure if such a long time is giving hints that there is a problem. I'm running Pilon using multi-thread with 45 processors.
Hello Pilon users,
I have already used Pilon to improve a genome assembly (400Mbp) using .bam alignments of a RADseq library. When I ran the program it took around 12h to finished the entire correction. However, I'm using Pilon again to improve another (shorter 273 Mbp) DeNovo genome and the program has been running for more than 57h. Is it normal? I'm not sure if such a long time is giving hints that there is a problem. I'm running Pilon using multi-thread with 45 processors.
Thanks in advance for any help you can provide