gmod / apollo3server Goto Github PK

move database pieces to neo4j

• preload neo4j "database"

https://hub.docker.com/repository/docker/gmod/apollo3server

Should be:

 grep -v ^#  loaded-data/yeast/raw/yeast.gff | cut -f3  | sort | uniq -c 
6913 CDS
7416 exon
   4 five_prime_UTR
7024 gene
6600 mRNA
  18 ncRNA
  12 pseudogene
  12 pseudogenic_transcript
  24 rRNA
   6 snRNA
  77 snoRNA
 299 tRNA
  91 transposable_element

May need to provide a uniqueName for Organisms (maybe we should call these "Genome's?")

https://docs.spring.io/spring-boot/docs/current/reference/html/spring-boot-features.html#boot-features-external-config

https://www.bookstack.cn/read/grails-v4.0/spilt.1.d6622541695e6abe.md

Orthologs/Paralog feature, orthomcl clustering of orthologs can be utilized to transfer annotation across organisms.

(a script tool using web services, but somehow automated)

If a terminator or Shine Dalgarno is associated with a gene, this should be indicated

In this case they would get coded similar to an exon, but not included in ORF calculations as they are usually upstream of the mRNA

• fix excessive # of genes
• fix out of order isoforms
• fix bad gene count (1400, should be 42K)

From apollo-performance code

grep "WormBase\tgene\t" loaded-data/worm/raw/worm.gff | wc -l
   44982

• create MRNAService and move relevant methods
• recast Transcript to MRNA
• move remaining exon methods to TranscriptServices or ExonServices
• create more methods and get rid of large if blocks / loops)
• convert methods to exclusively use MRNA instead of Transcript
• handle gene parent layer explicitly (if passed through)

• set test to unskip in unit test
• implement VCF output

https://travis-ci.org/github/GMOD/Apollo3Server/jobs/721555556

• Exception: Apollo failed to prepare the GFF3 file for download: {'error': 'Proxy for [org.bbop.apollo.attributes.Status] for association [org.bbop.apollo.attributes.Status] could not be initialized'}
• mkdir: cannot create directory ‘/data’: Permission denied
• EEEEE.SE............SSS.S..................................
  . .

From #22

• IO_test.test_export_fa_cds
• IO_test.test_export_fa_cdna
• IO_test.test_export_fa_peptide

replace "gormInstance instanceof Class" with "gormInstance.instanceOf(Class.class)"

https://gorm.grails.org/latest/neo4j/api/grails/gorm/api/GormInstanceOperations.html#instanceOf(D,%20java.lang.Class)

i.e.,

feature instanceof Gene

should be:

feature.instanceOf(Gene.class)

seems to be an issue with the organism test failing in the other environment

Not sure what the source of this is, as other dual exons work fine. Likely possibilities are:

1 - a casting error
2 - an error to do with the "executeQuery" environment versus the GORM layer
3 - should move to using criteria queries where possible as it is still GORM, but allows for greater flexibility

remove skipped test

Will not work as it can't parse the version.

the solution is to use java 8

Grails not found using grailsw
Using gradlew

FAILURE: Build failed with an exception.

* What went wrong:
Could not determine java version from '11.0.8'.

* Try:
Run with --stacktrace option to get the stack trace. Run with --info or --debug option to get more log output.
Gradle version:
Node Version: 10
Npm Version: 6
Yarn Version: 1
javac 1.8.0_265 found
javac installed
JDK 1.8 found: javac 1.8.0_265

FAILURE: Build failed with an exception.

* What went wrong:
Could not determine java version from '11.0.8'.

https://travis-ci.org/github/GMOD/Apollo3Server

• support embedding by default (replace default config with embedded one)
• get war file to build and run #26
• make sure that docker file copies over that one properly
• add proper docker file with env, etc. to support most environmental variables

https://hub.docker.com/repository/docker/gmod/apollo3server

https://guides.grails.org/grails-as-docker-container/guide/index.html

......S.............SSS.S..................................
----------------------------------------------------------------------
Ran 59 tests in 118.577s

OK (SKIP=5)

• IOTest.test_export_gff3

• fix overlap calculation when adding isoforms
• verify proper INITIAL creation of the CDS
• determine why it creates a single CDS model has the incorrect CDS that contains everything (should be 13096-13230) even though the calculation is correct

• verify that the isoform test works properly

arrow annotations load_gff3 --timing --batch_size=1 test_organism test-data/mrna-isoforms.gff

 nosetests test/annotations_test.py:AnnotationsTest.test_isoforms

http://jbrowse.org/jb2/docs/bcc2020_plugin_development/

The use case for this functionality is to allow curators/admin staff the ability to check users annotation and verify that it is accurate to their best assessment, putting a status for approval/rejected/pending to already status tagged "finished" models.

For some genomes in VEuPathDB we want to have some level of curatorial oversight especially with the highly curated organisms, so a way to approve a users annotation (mainly functional) would be beneficial. The way I thought this could be achieved is with an additional status that could be attributed to models that only an admin or role with higher permissions could use.

Example Joe Bloggs goes in drags a model to user created track, adds some GO terms and a product description, adds in the metadata and adds the status finished. A curator then goes in, looks at the report of all annotations with the finished status and goes through each annotation, either accepting the annotation as good or incorrect and adds another higher level status to approve/reject it....models with the admin tag are then parsed out and integrated.

• integrate https://github.com/steamcleaner/swagger-grails as a dependency
• build https://www.npmjs.com/package/swagger-ui-react into library app for consuming (and displaying) as part of apollo-gmod-js

Sync against this model: https://github.com/biolink/biolink-model/blob/master/biolink-model.yaml

Graphql references:

• https://github.com/biolink/biolink-model/blob/master/biolink-model.graphql#L687

• https://biolink.github.io/biolink-model/docs/Gene.html

• generate JSON Feature to reflect the Apollo model

• describe tree

• upload biolink schema

• do comparison

• Create a PR from d5f0dbf
• Integrate changes from here: 038b637

Requirements:

• consistent error handling (codes) and JSON return if more information needed
• . . . .

• Can swagger provide authentication layer?
• How does user information get loaded?

use OAuth2 . . . in resources.groovy

https://swagger.io/docs/specification/authentication/oauth2/

https://swagger.io/docs/specification/authentication/
. . . with an ApiKey . . or others

• multiple authentication schemas: swagger-api/swagger-ui#2399
• authentication: https://bluesliverx.github.io/grails-spring-security-oauth2-provider/v3/manual/guide/gettingStarted.html#createDomainClasses

This would be a potential way to help enforce Schem as well. Something to explroe

https://neo4j.com/developer/graphql/

. . . also: https://people.groupon.com/2019/swagql-use-graphql-in-minutes-leave-the-rest-to-us/