NanoSV performance issue

[Patricie34]

Hi,

I am trying to run NanoSV on my ONT data, but despite multiple attempts, the process takes an extremely long time, sometimes even several days, without producing any results.

I suspect it fails due to memory constraints. Currently, I'm running it with 8 threads and allocating 300 GB of RAM, hoping this would help. It has been running for more than a day now, but since no logs are being generated, I'm unsure whether it will finish successfully. Monitoring shows that the actual usage is much lower than requested (approximately 1 CPU and 50 GB of RAM).

The command I am using is:

NanoSV -t 8 -s /usr/bin/samtools -b ${bed} -o ${sample.name}.nanosv.vcf ${bam}

The bed file I'm using for hg38 contains approximately 400,000 positions.

I’ve read through other reported issues and saw that running NanoSV per chromosome is recommended for better performance. However, in our case, we need to detect interchromosomal structural variants, so this approach isn’t suitable.

Do you have any suggestions on how to speed up the process or identify what might be going wrong?

Thank you for your help.

Best regards,

Patricie