nf-core/phaseimpute

A bioinformatics pipeline to phase and impute genetic data

genomicsgenotypeimputationlow-pass-sequencingphasing

An incompatibility advisory with severity high has been issued for this version of the pipeline.
See the advisory entry for more information.

Launch version 1.0.0 https://github.com/nf-core/phaseimpute

Version history

Download .zip Download .tar.gz View on GitHub

Initial release of nf-core/phaseimpute, created with the nf-core template. Special thanks to Matthias Hörtenhuber, Mazzalab and Sofia Stamouli for the review of this release.

`Added`

#20 - Added automatic detection of vcf contigs for the reference panel and automatic renaming available.
#22 - Add validation step for concordance analysis. Input channels changed to match inputs steps. Outdir folder organised by steps. Modules config by subworkflows.
#26 - Added QUILT method.
#47 - Add possibility to remove samples from reference panel. Add glimpse2 chunking method. Add full-size test parameters.
#58 - Add external params posfile and chunks. Add glimpse2 phasing and imputation.
#67 - Export CSVs from each step.
#71 - Allow external panel to be used in step impute.
#97 - Add dog reference panel and config to test pipeline with other species.
#102 - Add dog panel test.
#119 - Add dog test with panelprep and imputation.
#118 - Explain how to customize arguments in the pipeline.
#111 - Add nf-test for all subworkflow, workflow, modules and functions.
#131 - Set normalisation as optional. Fix extension detection function. Add support for validation with vcf files. Concatenate vcf only if more than one file. Change --phased to --phase for consistency.
#143 - Improve contigs warning and error logging. The number of chromosomes contigs is summarized if above max_chr_names.
#146 - Add seed parameter for QUILT.
#164 - Add additional requirement on input schema "uniqueEntries": ["panel", "chr"] and end should be greater than start in regions.

`Changed`

#18 - Maps and region by chromosome. Update tests config files. Correct meta map propagation. test_impute and test_sim works.
#19 - Changed reference panel to accept a csv, update modules and subworkflows (glimpse1/2 and shapeit5)
#40 - Add STITCH method. Reorganize panelprep subworkflows.
#51 - Update all process and fix linting errors. Remove FASTQC added by the template.
#56 - Move to nf-test to check the output files names generated. Fix validation and concatenation by chromosomes missing. Add dedicated GLIMPSE1 subworkflow. Fix posfile generation to be done once for glimpse and stitch.
#68 - QUILT can handle external params chunks and hap-legend files.
#78 - Separate validate step from panel preparation.
#84 - Change depth computation to use SAMTOOLS_DEPTH and make separation by chromosome only if regions are specified.
#85 - Use external params in individual tests for tools.
#86 - Move BCFTOOLS_CONVERT to VCF_SITES_EXTRACT_BCFTOOLS.
#88 - Improve multiQC report with more information.
#91 - Update metro map with all steps and remove deprecated ones.
#93 - Add support for CRAM file.
#93 - Check contigs name at workflow level for BAM and VCF.
#93 - Samples remove with multi-allelics records.
#93 - Samtools merge in BAM_REGION subworkflow.
#93 - Fix glimpse2_phase output file names.
#93 - Fix fai combination to fasta.
#96 - Simplify csv export
#96 - Use only legend file as posfile for all imputation workflow.
#100 - Update bcftools, samtools, … nf-core modules. All indexing is now done with the file creation for most of them.
#101 - Set --compute_freq as false by default.
#102 - Compute chr name from whole vcf.
#102 - Only warn the user if some contigs are absent from files, the regions to compute is now the intersection of regions, panel, posfile, chunks, map.
#102 - Update all test and recompute snapshot to match new version of the phaseimpute test dataset.
#103 - Update GLIMPSE2_PHASE, GUNZIP and MULTIQC
#135 - Impute by batch of 100 individuals by default using --batch_size parameter. All individuals BAM files are gathered and VCF are allowed for GLIMPSE1 and GLIMPSE2. Channel preprocessing of stitch is done in stitch subworkflow. Genotype likelihood computation for GLIMPSE1 is now done outside of the subworkflow and merge the resulting vcf with all the samples. New test added to check batch separation. Improve usage.md documentation. Add validation to initialization of the pipeline to ensure compatibility between tools, steps and the files provided by the user.
#139 - Update all nf-core modules.
#146 - Remove conda CI check for PR due to Nextflow error.
#144 - Documentation updates.
#148 - Fix AWS fulltest github action for manual dispatch.
#149 - Remove the map file from the AWS fulltest.
#152 - Fix URLs in the documentation and remove tools citation in the README, use a white background for all images in the documentation.
#153 - Update and simplify subworkflows snapshot and check only for files names (no md5sum for bam and vcf files due to timestamp).
#157 - Add chunk_model as parameter for better control over GLIMPSE2_CHUNK and set window size in GLIMPSE1_CHUNK and GLIMPSE2_chunk to 4mb to reduce number of chunks (empirical).
#160 - Improve CHANGELOG.md and add details to usage.md
#158 - Remove frequency computation and phasing from full test to reduce cost and computational time.
#164 - Rename BAM_REGION_SAMTOOLS to BAM_EXTRACT_REGION_SAMTOOLS. Remove GLIMPSE2_SPLITREFERENCE as it is not used. Add more steps to test_all profile for more exhaustivity.
#163 - Improve configuration for demanding processes. Use Genome in a Bottle VCF benchmarking file for AWS full test. Moved from glimpse1 to glimpse2 for the full test profile.
#165 - Update metro map and add logo to the documentation.

`Fixed`

#15 - Changed test csv files to point to nf-core repository.
#16 - Removed outdir from test config files.
#65 - Separate stitch output by individuals.
#75 - Set frequency computation with VCFFIXUP process as optional with --compute_freq. Use GLIMPSE_CHUNK on panel vcf to compute the chunk and not makewindows on fasta.
#117 - Fix directories in CSV.
#151 - Fix Type not supported: class org.codehaus.groovy.runtime.GStringImpl error due to String test in getFileExtension().
#158 - Fix contigs usage when regions is only a subset of the given contigs (e.g. if panel file has the 22 chr and the region file only 2 then only the 2 common will be processed). Fix multiQC samples names for better comprehension. Fix -resume errors when ch_fasta is use by adding cache = 'lenient' in necessary processes. Fix --window-size of GLIMPSE_CHUNK from 4 to 4000000.
#153 - Fix getFileExtension function. Fix image in usage.md. Fix small warnings and errors with updated language server. def has been added when necessary, : use instead of , in assertions, _ added to variables not used in closures, for loop replaced by .each{}, remove unused code / input.
#161 - Fix VCF_SPLIT_BCFTOOLS when only one sample present by updating BCFTOOLS_PLUGINSPLIT and adding BCFTOOLS_QUERY to get truth samples names for renaming the resulting files.
#162 - Fix fai usage when provided by genomes parameter.
#164 - Improve documentation writing
#163 - Fix MULTIQC samples names (add post-processing for clean up FILTER_CHR_DWN, FILTER_CHR_INP, GAWK_ERROR_SPL, GAWK_RSQUARE_SPL). Fix output panel publisDir. Fix java version to 17 in ci.yml due to new nextflow version.

`Dependencies`

Dependency	New version
`bcftools`	1.20
`bedtools`	2.31.1
`gawk`	5.3.0
`glimpse-bio`	1.1.1
`glimpse-bio`	2.0.1
`gunzip`	1.10
`htslib`	1.21
`multiqc`	1.25.1
`r-quilt`	1.0.5
`r-stitch`	1.6.10
`samtools`	1.21
`shapeit5`	1.0.0
`tabix`	1.11
`vcflib`	1.0.3