unbound variable error

[drzhangyin]

Hello, I am new to PGScatalog pipeline, and I am trying to use it to calculate PRS strictly following the instruction.

When I tested nextflow run pgscatalog/pgsc_calc -profile test,singularity, I received the following error message:
Command error:
bash: line 15: __lmod_vx: unbound variable
.command.sh: line 2: 37 Illegal instruction combine_scorefiles -s PGS001229_22.txt -t GRCh37 -o scorefiles.txt.gz -l log_scorefiles.json -v

When I tried to calculate PRS using PGS000004 as an example, I received similar error message:
Command error:
bash: line 15: __lmod_vx: unbound variable
.command.sh: line 2: 36 Illegal instruction download_scorefiles -i PGS000004 -b GRCh37 -o $PWD -v -c pgsc_calc/1.3.2

It would be great to have your suggestions on how I can address this error.

Thanks!!

[nebfield]

Hello,

This looks like a problem with your local computing environment related to https://github.com/TACC/Lmod

Does $ nextflow run hello run OK?

You should see the following output:

Bonjour world!

Hello world!

Hola world!

Ciao world!

Also, please report the version of singularity you're running:

$ singularity --version
singularity version 3.8.7-1.el8

Thanks

[drzhangyin]

Thanks for the feedback!
The $ nextflow run hello runs well, and I see one of those four outputs each time.
The version of singularity I am running is 3.8.3-1.el7

[nebfield]

Could you update the version of lmod installed on your cluster?

TACC/Lmod#384

Or try overwriting the __lmod_vx environment variable as described here:

NOAA-EMC/hpc-stack#54 (comment)

[drzhangyin]

I will contact the cluster administrators to see if they can do so. Thank you~

[drzhangyin]

Hello and thanks again!
I retested $ nextflow run pgscatalog/pgsc_calc -profile test,singularity on a new cluster partition that with the latest version of lmod installed, and it ran with no error message. The output log file said "Pipeline completed successfully-". However, I didn't see score/ or match/ in the results directory, and the only folder I can see there is pipeline_info. Not sure is this normal?

I then tried to calculate PGS000790 and got the error message below. Is this solely due to issues with PGS Catalog API or could there be other reasons?

Command exit status:
1

Command output:
Error Connecting: HTTPSConnectionPool(host='www.pgscatalog.org', port=443): Max retries exceeded with url: /rest/score/search?pgs_ids=PGS000790 (Caused by NewConnectionError('<urllib3.connection.HTTPSConnection object at 0x2b78bb450700>: Failed to establish a new connection: [Errno 101] Network is unreachable'))

Command error:
bash: line 15: __lmod_vx: unbound variable
root: 2023-05-16 21:03:46 DEBUG Verbose logging enabled
pgscatalog_utils.download.download_scorefile: 2023-05-16 21:03:46 CRITICAL Downloading harmonized scoring file(s) in build: GRCh37.
pgscatalog_utils.download.download_scorefile: 2023-05-16 21:03:46 DEBUG --id set, querying scores
urllib3.connectionpool: 2023-05-16 21:03:46 DEBUG Starting new HTTPS connection (1): www.pgscatalog.org:443
pgscatalog_utils.download.score: 2023-05-16 21:05:53 WARNING No results in response from PGS Catalog API. Please check the PGS IDs.
pgscatalog_utils.download.score: 2023-05-16 21:05:53 ERROR Bad response from PGS Catalog API. Is ['PGS000790'] a valid ID?
Error Connecting: HTTPSConnectionPool(host='www.pgscatalog.org', port=443): Max retries exceeded with url: /rest/score/search?pgs_ids=PGS000790 (Caused by NewConnectionError('<urllib3.connection.HTTPSConnection object at 0x2b78bb450700>: Failed to establish a new connection: [Errno 101] Network is unreachable'))

[nebfield]

The results directory isn't normal. Running v1.3.2 on the test profile should output the following structure:

$ tree results/
results
├── match
│   ├── cineca-synthetic-subset_22_additive_0.scorefile.gz
│   ├── cineca-synthetic-subset_log.csv.gz
│   ├── cineca-synthetic-subset_summary.csv
│   └── versions.yml
├── pipeline_info
│   ├── execution_report_2023-05-17_08-48-13.html
│   ├── execution_timeline_2023-05-17_08-48-13.html
│   ├── execution_trace_2023-05-17_08-48-13.txt
│   └── pipeline_dag_2023-05-17_08-48-13.html
└── score
    ├── aggregated_scores.txt.gz
    ├── report.html
    └── versions.yml

You could try changing the output directory (--outdir) to a different place, maybe $HOME/results:

$ nextflow run pgscatalog/pgsc_calc -profile test,singularity --outdir $HOME/results

You could test re-running the workflow with the -resume option to see if your network problem was intermittent. The PGS Catalog API is OK:

$ curl -X GET "http://www.pgscatalog.org/rest/score/search?pgs_ids=PGS000790"
{"size":1,"count":1,"next":null,"previous":null,"results":[{"id":"PGS000790","name":"CC_Melanoma_IV","ftp_scoring_file":"https://ftp.ebi.ac.uk/pub/databases/spot/pgs/scores/PGS000790/ScoringFiles/PGS000790.txt.gz","ftp_harmonized_scoring_files":{"GRCh37":{"positions":"https://ftp.ebi.ac.uk/pub/databases/spot/pgs/scores/PGS000790/ScoringFiles/Harmonized/PGS000790_hmPOS_GRCh37.txt.gz"},"GRCh38":{"positions":"https://ftp.ebi.ac.uk/pub/databases/spot/pgs/scores/PGS000790/ScoringFiles/Harmonized/PGS000790_hmPOS_GRCh38.txt.gz"}},"publication":{"id":"PGP000186","title":"Pan-cancer analysis demonstrates that integrating polygenic risk scores with modifiable risk factors improves risk prediction.","doi":"10.1038/s41467-020-19600-4","PMID":33247094,"journal":"Nat Commun","firstauthor":"Kachuri L","date_publication":"2020-11-27"},"matches_publication":true,"samples_variants":[{"sample_number":2830,"sample_cases":null,"sample_controls":null,"sample_percent_male":null,"sample_age":null,"phenotyping_free":null,"followup_time":null,"ancestry_broad":"European","ancestry_free":null,"ancestry_country":"U.S.","ancestry_additional":null,"source_GWAS_catalog":"GCST001245","source_PMID":21926416,"source_DOI":null,"cohorts":[{"name_short":"MDACCS","name_full":"MD Anderson Cancer Center Study","name_others":null}],"cohorts_additional":null},{"sample_number":1728,"sample_cases":null,"sample_controls":null,"sample_percent_male":null,"sample_age":null,"phenotyping_free":null,"followup_time":null,"ancestry_broad":"European","ancestry_free":null,"ancestry_country":"Australia","ancestry_additional":null,"source_GWAS_catalog":"GCST000198","source_PMID":18488026,"source_DOI":null,"cohorts":[{"name_short":"Q-MEGA","name_full":"Queensland study of Melanoma: Environment and Genetic Associations","name_others":null}],"cohorts_additional":null},{"sample_number":36077,"sample_cases":null,"sample_controls":null,"sample_percent_male":null,"sample_age":null,"phenotyping_free":null,"followup_time":null,"ancestry_broad":"European","ancestry_free":null,"ancestry_country":"U.S., Israel, Australia, Germany, U.K., France","ancestry_additional":null,"source_GWAS_catalog":"GCST003061","source_PMID":26237428,"source_DOI":null,"cohorts":[],"cohorts_additional":null},{"sample_number":6555,"sample_cases":null,"sample_controls":null,"sample_percent_male":null,"sample_age":null,"phenotyping_free":null,"followup_time":null,"ancestry_broad":"European","ancestry_free":null,"ancestry_country":"Australia","ancestry_additional":null,"source_GWAS_catalog":"GCST001266","source_PMID":21983785,"source_DOI":null,"cohorts":[{"name_short":"QIMR","name_full":"Queensland Institute of Medical Research","name_others":null},{"name_short":"BATS","name_full":"Brisbane Adolescent Twins Study","name_others":null},{"name_short":"AMFS","name_full":"Australian Melanoma Family Study","name_others":null},{"name_short":"Q-MEGA","name_full":"Queensland study of Melanoma: Environment and Genetic Associations","name_others":null}],"cohorts_additional":null},{"sample_number":291407,"sample_cases":null,"sample_controls":null,"sample_percent_male":null,"sample_age":null,"phenotyping_free":null,"followup_time":null,"ancestry_broad":"European","ancestry_free":null,"ancestry_country":"U.S.","ancestry_additional":null,"source_GWAS_catalog":"GCST004142","source_PMID":28212542,"source_DOI":null,"cohorts":[{"name_short":"23andMe","name_full":"23andMe","name_others":null}],"cohorts_additional":null},{"sample_number":6122,"sample_cases":null,"sample_controls":null,"sample_percent_male":null,"sample_age":null,"phenotyping_free":null,"followup_time":null,"ancestry_broad":"European","ancestry_free":null,"ancestry_country":"U.S.","ancestry_additional":null,"source_GWAS_catalog":"GCST002514","source_PMID":24980573,"source_DOI":null,"cohorts":[{"name_short":"NHS","name_full":"Nurses Health Study","name_others":null},{"name_short":"HPFS","name_full":"Health Professional Follow-up Study","name_others":null}],"cohorts_additional":null},{"sample_number":10422,"sample_cases":null,"sample_controls":null,"sample_percent_male":null,"sample_age":null,"phenotyping_free":null,"followup_time":null,"ancestry_broad":"European","ancestry_free":null,"ancestry_country":"Sweden,U.S.,Poland,Israel,Australia,Italy,Netherlands,U.K.,Spain,France,Norway","ancestry_additional":null,"source_GWAS_catalog":"GCST001267","source_PMID":21983787,"source_DOI":null,"cohorts":[{"name_short":"GenoMEL","name_full":"GenoMEL","name_others":null},{"name_short":"B58C","name_full":"British 1958 Birth Cohort Study","name_others":null},{"name_short":"UKBS","name_full":"UK National Blood Donor Service","name_others":null},{"name_short":"CNG","name_full":"Centre National de G_notypage","name_others":null}],"cohorts_additional":null},{"sample_number":5456,"sample_cases":null,"sample_controls":null,"sample_percent_male":null,"sample_age":null,"phenotyping_free":null,"followup_time":null,"ancestry_broad":"European","ancestry_free":null,"ancestry_country":"Sweden,Australia,Italy,Netherlands,U.K.,Spain,France","ancestry_additional":null,"source_GWAS_catalog":"GCST000437","source_PMID":19578364,"source_DOI":null,"cohorts":[{"name_short":"GenoMEL","name_full":"GenoMEL","name_others":null},{"name_short":"WTCCC","name_full":"Wellcome Trust Case Control Consortium","name_others":null},{"name_short":"CNG","name_full":"Centre National de G_notypage","name_others":null}],"cohorts_additional":null},{"sample_number":4919,"sample_cases":null,"sample_controls":null,"sample_percent_male":null,"sample_age":null,"phenotyping_free":null,"followup_time":null,"ancestry_broad":"European","ancestry_free":null,"ancestry_country":null,"ancestry_additional":null,"source_GWAS_catalog":"GCST001886","source_PMID":23455637,"source_DOI":null,"cohorts":[{"name_short":"GenoMEL","name_full":"GenoMEL","name_others":null},{"name_short":"WTCCC","name_full":"Wellcome Trust Case Control Consortium","name_others":null}],"cohorts_additional":null}],"samples_training":[],"trait_reported":"Melanoma","trait_additional":null,"trait_efo":[{"id":"EFO_0000756","label":"melanoma","description":"A malignant, usually aggressive tumor composed of atypical, neoplastic melanocytes. Most often, melanomas arise in the skin (cutaneous melanomas) and include the following histologic subtypes: superficial spreading melanoma, nodular melanoma, acral lentiginous melanoma, and lentigo maligna melanoma. Cutaneous melanomas may arise from acquired or congenital melanocytic or dysplastic nevi. Melanomas may also arise in other anatomic sites including the gastrointestinal system, eye, urinary tract, and reproductive system. Melanomas frequently metastasize to lymph nodes, liver, lungs, and brain. [NCIT: C3224]","url":"http://www.ebi.ac.uk/efo/EFO_0000756"}],"method_name":"24 variants from Graff et al (PGS000079) with inverse variant weights","method_params":"NR","variants_number":24,"variants_interactions":0,"variants_genomebuild":"GRCh37","weight_type":"Inverse-variance weighting","ancestry_distribution":{"eval":{"dist":{"EUR":100},"count":1},"gwas":{"dist":{"EUR":100},"count":365516}},"date_release":"2021-05-28","license":"PGS obtained from the Catalog should be cited appropriately, and used in accordance with any licensing restrictions set by the authors. See EBI Terms of Use (https://www.ebi.ac.uk/about/terms-of-use/) for additional details."}]}%

If you have consistent problems with the PGS Catalog API, maybe your system administrator could help to fix your network problems.

I hope that helps 😄

[drzhangyin]

Thanks, and this is very helpful. 😄
Just a quick update that I retested $ nextflow run pgscatalog/pgsc_calc -profile test,singularity using another output directory and it now works.
The network problem seems to be consistent, so I am contacting the cluster administrators to see if they can help me fix it~