ALIGNER:chainMerge (chainmerge) FAILED about nf-lo

[RenzoTale88]

Hi @shwetaramdas
thanks for using nf-LO.
Could you try running the workflow with decompressed fasta files? I believe that might be the source of the issue.
Thanks
Andrea

[shwetaramdas]

Dear Andrea,

I'm trying this now and it has been running for a while--I will update you about whether it works. Thank you!

[shwetaramdas]

Dear Andrea,

I tried with unzipped input files, and ran into the same error. Could you please help me figure this out?

[a4/840c6e] process > ALIGNER:lastz (lastz_med.095.tgt183) [100%] 350924 of 350924 ✔
[8f/0ac640] process > ALIGNER:axtChain (axtchain_m) [100%] 350924 of 350924 ✔
[fb/7305fc] process > ALIGNER:chainMerge (chainmerge) [100%] 3 of 3, failed: 3, retries: 2 ✔
[- ] process > ALIGNER:chainNet -
[- ] process > ALIGNER:netSynt -
[- ] process > ALIGNER:chainsubset -
[- ] process > ALIGNER:chain2maf -
[- ] process > ALIGNER:name_maf_seq -
[- ] process > ALIGNER:mafstats -

Thank you,
Shweta

[RenzoTale88]

@shwetaramdas of course, you can try figure out the issue by looking into the folder starting with ./work/fb/7305fc (e.g. by using ls -al ./work/fb/7305fc*/). This folder should contain many files, including one called .command.err, that should contain some more info about why the work failed (cat ./work/fb/7305fc*/.command.err).
One more suggestion is to tweak the configuration for your genome. I can see that you have processed over 350K fragments, which is way too mamy. Try to increase the fragment size (--tgtSize and --srcSize) and the chunk overlap (--tgtSize and --srcSize). You can find some examples with large genomes in the supplementary table 1 of the nf-LO manuscript. I'll update the documentation to discuss about avoiding over fragmentation since I realise it is not well documented.

[shwetaramdas]

Thank you, @RenzoTale88 . I will try that. As an alternative, I used default parameters. and split the source files by chromosome (and kept the target fasta file intact). Those ran without error. Would these chains (one per source chromosome) be reasonable to use as well?

Thank you,
Shweta

[RenzoTale88]

I would recommend to try editing the parameter, in order to consider all possible combination of chromosomes. The default parameters are ok for small genomes, but for larger ones I would definitely recommend larger fragments. Also, the fact that you can run it chromosome by chromosome means that it is likely an issue of overfragmentation. Andrea

…

________________________________ From: Shweta ***@***.***> Sent: Friday, February 10, 2023 7:22:13 AM To: evotools/nf-LO ***@***.***> Cc: Andrea Talenti ***@***.***>; Mention ***@***.***> Subject: Re: [evotools/nf-LO] ALIGNER:chainMerge (chainmerge) FAILED (Issue #11) Thank you, @RenzoTale88<https://github.com/RenzoTale88> . I will try that. As an alternative, I used default parameters. and split the source files by chromosome (and kept the target fasta file intact). Those ran without error. Would these chains (one per source chromosome) be reasonable to use as well? Thank you, Shweta — Reply to this email directly, view it on GitHub<#11 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AFRTPKE4VDN4UWXEB6U7JX3WWXUCLANCNFSM6AAAAAAUSG67IE>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[RenzoTale88]

Closing