Add Transvar - #8250
Open
joachimwolff wants to merge 6 commits into
Open
Conversation
TransVar annotates genomic, cDNA and protein variants against one or more transcript sets. This adds the tool wrapper plus a data manager that puts its transcript annotation databases in a Galaxy managed location. Wrapper: - Single variants typed into the form are passed with "-i". A positional query is not valid TransVar syntax, and the default text sanitizer replaces the ">" of an HGVS substitution, so the valid character set is spelled out for the query parameter. - VCF input is offered only in genomic mode, the only mode that accepts it, and produces VCF output. - A minimal generated config is passed through TRANSVAR_CFG. TransVar looks up a section for the reference version before reading the command line and aborts if it is missing, and this keeps it away from the config inside the installed Python package. - The reference genome comes from fasta_indexes or the history. TransVar needs it to resolve codons; without it every query returns an empty table. - version_command tolerates "transvar --version" exiting 1, which otherwise aborts the job before the tool command runs. Data manager: - Downloads the pre-built .transvardb files into a Galaxy managed directory instead of the installed Python package directory, and registers them in the transvar_annotations data table. - URLs come from the installed TransVar release, so they follow the requirement version. Files are renamed after the command line flag that consumes them, so the wrapper needs no knowledge of TransVar's internal file naming. - Offers only openly licensed, redistributable annotation sets: RefSeq, CCDS, Ensembl, GENCODE and the UCSC RefGene and knownGene tables. AceView is excluded for lack of a stated redistribution licence, and the genome and dbSNP downloads are left to Galaxy's own reference data management. - mm9 is not offered: TransVar lists it, but no pre-built databases are published for it. Tested with planemo against release_24.2: 7/7 tool tests and 1/1 data manager test pass. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…ment Galaxy reduces a job's environment to HOME, LC_CTYPE, PATH and TMPDIR (galaxy.util.commands.new_clean_env), so a proxy configured for the Galaxy server itself never reaches this tool and the download fails with a connection timeout. Document how to set it on the job destination and say so in the error message, since the failure otherwise gives no hint. Also make the download survive a hiccup: a full annotation set is around 35 files and a few hundred megabytes from a single academic server, and a single failure part way through used to abort the whole run. Retry up to three times with backoff and a socket timeout, but do not retry a 4xx, which is how an absent optional idmap index is detected. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Galaxy merges the fields of several data table rows that share a value into one comma-joined string, so a stale or duplicated transvar_annotations entry produced a command line with --refversion 'hg19,hg19' and a database path made of two paths glued together. TransVar then fell back to resolving the path relative to TRANSVAR_DOWNLOAD_DIR and crashed inside posixpath.join with "expected str, bytes or os.PathLike object, not NoneType", which says nothing about the real problem. allow_duplicate_entries="False" does not help here: Galaxy hashes the whole row, so two rows with the same value but different paths both survive. Check each selected database file up front and name the missing source and the table it came from, and set TRANSVAR_DOWNLOAD_DIR so TransVar's own fallback cannot dereference None. Adds a test with a data table entry that claims a source it does not contain. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
github-project-automation
Bot
moved this to Triage/Discuss
in Galaxy Tool Developers SIG Board
The IUC lint job runs planemo with --biocontainers at fail-level warn, and requiring transvar together with samtools has no registered BioContainer, so linting failed with "Failed to find a BioContainer registered for these requirements". samtools was only there to index a reference FASTA taken from the history. Drop that option and read the genome from the fasta_indexes data table, which leaves transvar as the single requirement and its own BioContainer is found. Nothing is really lost: the transcript databases exist only for a handful of assemblies and have to be installed by an administrator with the data manager anyway, so the matching genome comes from the same hands, and pairing the two from curated lists makes it harder to annotate against the wrong assembly. Also replace the example proxy URL in the data manager help with a placeholder, since shed_lint --urls tries to fetch every URL it finds there. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Contributor
Author
|
please review |
tools-iuc
Bot
moved this from Triage/Discuss
to Needs Review
in Galaxy Tool Developers SIG Board
bgruening
reviewed
Jul 31, 2026
| <expand macro="requirements"/> | ||
| <!-- TransVar prints its version to stderr and then exits 1, which would | ||
| abort the job before the tool command runs. --> | ||
| <version_command>transvar --version 2>&1 || true</version_command> |
| ## If a database is missing, TransVar falls back to looking it up relative to | ||
| ## TRANSVAR_DOWNLOAD_DIR and crashes with an unreadable TypeError when that is | ||
| ## unset. Check up front so the cause is stated instead. This also catches a | ||
| ## transvar_annotations table holding several rows with the same value, which |
Member
There was a problem hiding this comment.
that is not true, you can specify on the data table that duplicates are not allowed.
Member
There was a problem hiding this comment.
Not sure if this entire thing here is needed or if its just made up by an LLM to work around another issue.
| #end if | ||
| ]]></command> | ||
| <configfiles> | ||
| <configfile name="transvar_cfg"><![CDATA[[DEFAULT] |
Member
There was a problem hiding this comment.
you could fill out this config file instead of using the parameters do you? Is that maybe easier, especially when you need to have the config eitherway?
Review feedback: the config file has to be generated anyway, so it can carry the reference and the database paths instead of repeating them on the command line. TransVar is built for this. --reference defaults to _DEF_ and a bare --ucsc/--refseq/... takes const=_DEF_, both of which mean "read the value from the config", so the command line only has to name which databases to switch on: transvar panno --refversion 'hg19' --ccds --refseq --ucsc -i 'KRAS:p.G12D' Also drop the shell loop that checked every database file up front. It was guarding against a data table holding several rows with the same value, which is not something the tool should paper over. Keeping TRANSVAR_DOWNLOAD_DIR set is enough: without it TransVar dies in posixpath.join with "expected str, bytes or os.PathLike object, not NoneType" when it cannot open a database, with it the same case reports a plain FileNotFoundError naming the path. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Contributor
Author
|
Thanks for the review, three of the four are addressed in the commit. || true on the version command: transvar --version prints the version to stderr and then exits 1, which is an upstream bug. Galaxy puts the version command ahead of the tool command in tool_script.sh under set -e, so without suppressing it the job dies before TransVar runs, every test failed with exit 1 and an empty output until I added it. The version string is captured correctly either way. I can switch it to the ( ... || true ) form used in tools/vapor if you prefer that style; there's no way to distinguish a real failure from the usual exit 1, since upstream always returns it. |
Contributor
Author
|
please review |
|
💬 There are unresolved review comments on this PR. Please resolve them before requesting a review. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This adds transvar with the wrapper and a datamanger to download lookup databases. TransVar is a multi-way annotator for genetic elements and genetic variations, see https://bioinformatics.mdanderson.org/public-software/transvar/ and https://github.com/zwdzwd/transvar and is under MIT license.
FOR CONTRIBUTOR:
There are two labels that allow to ignore specific (false positive) tool linter errors:
skip-version-check: Use it if only a subset of the tools has been updated in a suite.skip-url-check: Use it if github CI sees 403 errors, but the URLs work.To request a review once your PR is ready, comment "please review" on the PR. This will
automatically apply the
ready-for-reviewlabel if the PR is not a draft, all reviewthreads are resolved, and all CI checks have passed.