Closed AnotherSimon closed 6 years ago
Well certainly a pasa issue.... what it is trying to do here is not re-run PASA alignment step if it doesn't have to, so it looks into the DB to see if it has been run already, since it seems to think it has been run before, then it skips that step and tries to run PASA comparison (or update gene models).
If you ran PASA previously using a different --species
name this might be the issue. It is setup so that if you run funannotate train
followed by funanntoate predict
and then finally add UTRs with funannotate update
that it will re-use the PASA config file located in the training
subfolder. To bypass this, say perhaps PASA failed and you ran it with another config file successfully, you can pass that config file to funannotate update --pasa_config
.
So likely what is happening is that there is a pasa config file in the training
directory that is causing the error. So you can 1) delete it and it will re-run alignment step, or 2) pass the script the correct pasa_config file used to run PASA the first time.
deleting the config file (./My_bug/training_misc/pasa/alignAssembly.txt) did indeed lead to remapping but the same error pops up ("Table 'My_bug.annotation_updates' doesn't exist"). I found an earlier error in funannotate-update.log:
Processing CMD: 33/33 12:31:00 CMD: /home/simon/software/PASApipeline-pasa-v2.2.0/scripts/describe_alignment_assemblies_cgi_convert.dbi -M My_bug > My_bug.pasa_assemblies_described.txt
########################################################################## Finished. Please visit the Assembly and Annotation Comparison results at: /status_report.cgi?db=My_bug ##########################################################################
2018-02-21 12:31:32,219: ERROR 1071 (42000) at line 172 in file: '/home/simon/software/PASApipeline-pasa-v2.2.0/schema/cdna_alignment_mysqlschema': Specified key was too long; max key length is 1000 bytes Committing... CMD: /data1/home/simon/software/PASApipeline-pasa-v2.2.0/scripts/process_GMAP_alignments_gff3_chimeras_ok.pl --genome My_bug/update_misc/genome.fa --transcripts My_bug/training/trinity.fasta --CPU 12 -N 1 -I 3000 > gmap.spliced_alignments.gff3 ...
Oh okay. I've seen this before. https://github.com/PASApipeline/PASApipeline/issues/45. Change the key length in the mysql schema.
I applied the fix suggested in the referenced issue and reran the pipeline from scratch. The error manifests differently but it's still there:
##
Processing CMD:16:38:57 CMD: /home/simon/software/PASApipeline-pasa-v2.2.0/scripts/cDNA_annotation_comparer.dbi -G /data1/home/simon/IMB/TDlab/Annotation/funannot_test/My_bug/update_misc/genome.fa --CPU 12 -M My_bug > pasa_run.log.dir/My_bug.annotation_compare.38322.out
2018-02-26 16:39:04,400: Warning, no genes retrieved for contig_id: contig.25 ... Warning, no genes retrieved for contig_id: contig.138 DBD::mysql::st execute failed: Table 'My_bug.annotation_updates' doesn't exist at /home/simon/software/PASApipeline-pasa-v2.2.0/PerlLib//Mysql_connect.pm line 124. DBD::mysql::st execute failed: Table 'My_bug.annotation_updates' doesn't exist at /home/simon/software/PASApipeline-pasa-v2.2.0/PerlLib//Mysql_connect.pm line 124.
====
Failed query:Thread 2 terminated abnormally: Failed query:
Hmm, not sure where I went wrong but applying both the var char length fix and deleting the old PASA alignment config file seems to have done the trick.
Do you know if an update to PASA will be forthcoming or should this be included somewhere in the funannotate documentation?
I think this issue can be considered solved.
@AnotherSimon how did you fix the character length in Mysql. Googling this it implies you have to recompile mysql? https://stackoverflow.com/questions/5898518/how-do-i-increase-key-length-in-mysql-5-1
cheers, Pete
It's been a while but if I remember correctly, it went something along these lines: https://stackoverflow.com/questions/1279568/how-can-i-modify-the-size-of-column-in-a-mysql-table Just need to apply it to the correct table and column (this is where the memory gets fuzzy, see also Jon's comment pointing to PASA issue 45).
Thank you very much for your reply. I think my issue is that I am using the Docker image which doesn’t recognise mysql as a command. – I will sort it.
Cheers, Pete
From: AnotherSimon notifications@github.com Sent: 20 March 2019 08:46 To: nextgenusfs/funannotate funannotate@noreply.github.com Cc: Peter Thorpe pjt6@st-andrews.ac.uk; Comment comment@noreply.github.com Subject: {Disarmed} Re: [nextgenusfs/funannotate] Pasa issue in update step (#141)
It's been a while but if I remember correctly, it went something along these lines: https://stackoverflow.com/questions/1279568/how-can-i-modify-the-size-of-column-in-a-mysql-table Just need to apply it to the correct table and column (this is where the memory gets fuzzy, see also Jon's comment pointing to PASA issue 45).
— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/nextgenusfs/funannotate/issues/141#issuecomment-474738478, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AJhCqBMUeF34e30eKj-lKbf-pFCna1ybks5vYfU1gaJpZM4SLb0f. {"api_version":"1.0","publisher":{"api_key":"05dde50f1d1a384dd78767c55493e4bb","name":"GitHub"},"entity":{"external_key":"github/nextgenusfs/funannotate","title":"nextgenusfs/funannotate","subtitle":"GitHub repository","main_image_url":"https://github.githubassets.com/images/email/message_cards/header.png","avatar_image_url":"https://github.githubassets.com/images/email/message_cards/avatar.png","action":{"name":"Open in GitHub","url":"https://github.com/nextgenusfs/funannotate"}},"updates":{"snippets":[{"icon":"PERSON","message":"@AnotherSimon in #141: It's been a while but if I remember correctly, it went something along these lines:\r\nhttps://stackoverflow.com/questions/1279568/how-can-i-modify-the-size-of-column-in-a-mysql-table\r\nJust need to apply it to the correct table and column (this is where the memory gets fuzzy, see also Jon's comment pointing to PASA issue 45)."}],"action":{"name":"View Issue","url":"https://github.com/nextgenusfs/funannotate/issues/141#issuecomment-474738478"}}} [ { "@context": "http://schema.org", "@type": "EmailMessage", "potentialAction": { "@type": "ViewAction", "target": "https://github.com/nextgenusfs/funannotate/issues/141#issuecomment-474738478", "url": "https://github.com/nextgenusfs/funannotate/issues/141#issuecomment-474738478", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { "@type": "Organization", "name": "GitHub", "url": "https://github.com" } } ]
According to the manual, after running gene prediction (with RNAseq support), one should run update "to add UTR data to the predictions and fix gene models". When running:
funannotate update -i MyOrganism
The process seems to start off alright but soon fails:
Relevant logging info was found in funannotate-update.log:
Manually connecting to the mysql instance confirms that this table does not exist for db "My_bug_strainNum". Prediction step did not raise any errors or highly visible warnings.
PS: This may or may not be relevant but the first time running predict it failed because I needed to add the hooks to
@INC
as follows:My_bug_strainNum was manually dropped and prediction rerun.