<#2373 Add P. falciparum WGS and amplicon test dat...
# test-data
g
#2373 Add P. falciparum WGS and amplicon test datasets Pull request opened by vmduarut Summary Initial draft engineering fixture to exercise pathogenepidemiology and work towards reliable CI. Biological value and expected biological outputs have not been studied; no pipeline or CI pass is claimed. • Includes paired-end P. falciparum WGS-derived reads from ERR10375861, ERR10375871 and ERR10375886, paired Illumina amplicon reads from ERR13167700, single-file Nanopore amplicon reads from ERR13167482, a chromosome 1 FASTA and a BED interval. • The WGS reads were selected from chromosome 1 interval
Pf3D7_01_v3:185000–205000
(1-based inclusive; BED
Pf3D7_01_v3 184999 205000
). The interval includes PF3D7_0104300 (PfUBP1), a locus associated with artemisinin susceptibility. Mate fetching can include reads outside the interval. • Sequencing reads were sourced from SRA. Earlier checked ENA accession metadata complements the run and assay descriptions. The WGS README describes the files, provenance and how to generate them. Testing and limitations A dataset with two contig intervals was tested, but processing took too long in the MAFFT alignment stage. While we improve that stage, I propose temporarily deferring full multi-interval channel-logic testing and relying on stub tests for that logic. Full multi-interval integration coverage can be revisited once alignment performance improves. A follow-up PR in the nf-core/pathogenepidemiology pipeline is planned to test these datasets and work towards verifying CI. Samplesheets will be located under
assets/
in the nf-core/pathogenepidemiology pipeline repository. After merge, files will be available under
https://raw.githubusercontent.com/nf-core/test-datasets/pathogenepidemiology/data/plasmodium_falciparum
. nf-core/test-datasets
Review on #2373 Add P. falciparum WGS and amplicon test datasets Review comment by LouisLeNezet on vmduarut's pull request Hi, This first dataset seems a bit big for CI/CD but this highly depends on what is necessary for your tools to run. Using lighter files will greatly help you to reduce the computational load during development and help you iterate faster. I would suggest the following: • subsample your different files to a subregion of chr1 of 10-50kb, the aim would be to reduce the file size to just a few KB/MB. • add a second set of a different chromosome so that you can test that your channel manipulation works on multiple chromosome and stay in sync • add a script or markdown file describing the data generation (as done in the description of the PR) that you could run to generate back the data from scratch (this is highly useful in the long run). nf-core/test-datasets
Review on #2373 Add P. falciparum WGS and amplicon test datasets Review approved by LouisLeNezet on vmduarut's pull request This is indeed much lighter. nf-core/test-datasets
Review on #2373 Add P. falciparum WGS and amplicon test datasets Review approved by BilalsGituation on vmduarut's pull request nf-core/test-datasets