Memory issue OSError: [Errno 12] Cannot allocate memory

[CWYuan08]

Hi I am trying to correct some cDNA reads from a PrormethION run, the data is very big (120M reads!).
The isONcorrect pipeline was running and generated the full_length.fq but it stopped at clustering..
How much memory I should allocate for this job or should I split them into smaller batches?

Many many thanks!

Best,
CW

[ksahlin]

Hi @CWYuan08

That's a lot of reads! Either

1. try the C++ version of isONclust: https://github.com/nanoporetech/isONclust2 or
2. sort the reads according to length and split into many batches of like 5M reads each and apply isONclust individually on them.

I hope the C++ version will work though, as it does the sorting and batching for you.

[CWYuan08]

Dear @ksahlin,

many thanks! I will try isONclust2, I am actually trying to correct the reads, does isONclust2 still provide the same function?

Best,
CW

[ksahlin]

isONcorrect needs the clusters from isONclust(/2) to work efficiently, so in practice, yes.