# Segmentation fault (core dumped) happened when running step2_purge_dups

[guang-xi]

hi Dengfeng: something was wrong when i used purge_dup step 2, my paf file produced in step1 is about 1.3G, bellow is my command lines: ######################## /hwfssz1/ST_AGRIC/USER/chenhongyun/app/purge_dups/bin/purge_dups -2 -T cutoffs -c PB.base.cov genome.fasta.split.self.paf.gz > dups.bed ######################## and my output errror is ——/opt/gridengine/default/spool/cngb-compute-f21-4/job_scripts/3100583: line 1: 65185 Segmentation fault (core dumped)

Do you have some ideas or advice ,please?

[Tom-Jenkins]

Hi Dengfeng, I have the same error message as @guang-xi while trying to run the purge_dups command:

[M::main] finish parsing params [M::main] finish reading hits [M::main] finish reading cutoffs [M::main] finish reading coverages /var/spool/slurmd/job47295/slurm_script: line 22: 1945 Segmentation fault (core dumped) purge_dups -2 -T cutoffs -c PB.base.cov remap.paf > dups.bed 2> purge_dups.log

The commands I have used: minimap2 -c --cs -x map-pb --secondary=no assembly.fasta pacbio_reads.fq -t 16 > remap.paf pbcstat remap.paf hist_plot.py PB.stat coverage --xmax 150

I defined my own cutoff parameters based on the histogram:

calcuts -l 5 -m 25 -u 120 -d 1 PB.stat > cutoffs 2>calcults.log purge_dups -2 -T cutoffs -c PB.base.cov remap.paf > dups.bed

The specifications of the hardware I am using are:

• CentOS Linux, VERSION="7 (Core)"
• 2x Intel Haswell E5-2640v3 2.6GHz
• 16 cores
• 128GB memory

My remap.paf file is very large (31G) because I have a large genome and a lot of PacBio reads aligned to it. Any advice would be great, thanks!

[dfguan]

Sorry for the late reply, hope you have resolved the problem. If you still have, the input should not be remap.paf, instead it should be the splited ref-vs-split ref alignments, please refer to Readme step 1. Best, Dengfeng.