Husen Umer
01/13/2025, 6:08 AMincludeConfig '/internal/path/readable/by/all/genomes.config'
PS: I did ask a similar question earlier but didn't get to any conclusions so I am trying my luck again! thanks for your patienceJames Fellows Yates
01/13/2025, 7:46 AMJoon Klaps
01/13/2025, 7:49 AMdef CUSTOM_GENOMES_PATH = System.getenv("CUSTOM_GENOMES_PATH") ?: "/default/path/to/public/repo"
In the docs mention that they need to patch this ENV variable to the shared config file.pontus
01/13/2025, 8:32 AMAdam Talbot
01/13/2025, 10:20 AMAdam Talbot
01/13/2025, 10:20 AMparams.genomesAdam Talbot
01/13/2025, 10:21 AMparams {
genomes {
"mygenome" {
fasta = "path.fasta"
...Adam Talbot
01/13/2025, 10:22 AMMaxime Garcia
01/13/2025, 10:22 AMAdam Talbot
01/13/2025, 10:26 AMCurrently, I have added this line to the profile config but the lint test is failing during the pull request since the config file is not found in the configs repo.
Thing is, this makes sense because your pipeline can't run on Github. Indeed, no-one can run your pipeline outside of your home institution. Rather than making thinking about how to get around the genomes config, think about making it run everywhere. Can you include a custom trimmed down genome within the repo itself for testing? How about including yourincludeConfig '/internal/path/readable/by/all/genomes.config'
genomes.config file in the repo?pontus
01/13/2025, 10:40 AMHusen Umer
01/14/2025, 8:46 AMgenomes.config file is changing frequently by adding/updating genome files to it therefore I thought to just keep it somewhere local instead of pushing to the configs repo all the time!Husen Umer
01/14/2025, 8:47 AMJoon Klaps
01/14/2025, 8:47 AMnf-core/configs.Husen Umer
01/14/2025, 8:48 AMJoon Klaps
01/14/2025, 8:51 AMincludeConfig "<https://raw.githubusercontent.com/husensofteng/nf-kaust-config/path/to/custom/config>" think it can be done like that. Nothing to fancy :DHusen Umer
01/14/2025, 8:54 AMPhil Ewels
01/14/2025, 9:52 AMPhil Ewels
01/14/2025, 9:53 AMPhil Ewels
01/14/2025, 9:53 AMHusen Umer
01/14/2025, 10:28 AMRefGenie first, then moved to genomepy, and finally it seemed that just having a script to download fasta and gtf files from Ensembl and NCBI for a given species, then running the reference workflow that @Maxime Garcia is working on to generate indices, etc would be the best option to generate an internally maintained genome resource. Finally, a script scans the genome directory and generates this config to make the genome files accessible by the workflows 🫠Phil Ewels
01/14/2025, 10:37 AM