The format is based on Keep a Changelog and this project adheres to Semantic Versioning.
Patch release with minor improvements in docs and validation of the pipeline.
- [#218] - Improve the input validation of the pipeline. This will prevent multiple downstream errors
- [#221] - Update pipeline documentation for Shiny.
Nova Icaria is a beach in Barcelona. We started from Somorrostro (beach in front of CRG) and we are going up the way to Costa Brava.
Minor release of nf-core/multiplesequencealign.
This fixes few bugs, improves the documentation and adds the possibility to avoid the samplesheet and toolsheet and pass direclty the input files and tools directives.
Special thanks to the following for their contributions to the release:
Thank you to everyone else that has contributed by reporting bugs, enhancements or in any other way, shape or form.
- [#174] - Add the chaining of proteinfold output to MSA input.
- [#177] - Add MAFFT guidetree.
- [#179] - Add visualisation subworkflow and final csv merging onComplete.
- [#189] - Add GPU support and the alternatives to samplesheet and toolsheet (aligner, seqs)
- [#192] - Update documentation and fix bug on the merging of the
csvfileonComplete.
| Old parameter | New parameter |
|---|---|
--seqs |
|
--pdb_dir |
|
--aligner |
|
--args_aligner |
|
--tree |
|
--args_tree |
|
--skip_validation |
|
--skip_preprocessing |
Somorrostro is a beach in Barcelona.
Initial release of nf-core/multiplesequencealign, created with the nf-core template.
- [#29] - Add FAMSA_ALIGN module.
- [#38] - Add FAMSA_GUIDETREE module.
- [#42] - Add MAFFT module.
- [#51] - Add CLUSTALO module.
- [#41] - Add KALIGN module.
- [#48] - Integrate NF-VALIDATION PLUGIN.
- [#45] - Add module LearnMSA
- [#35] - Add module Tcoffee_align.
- [#60] - Add module Tcoffee3D_align and handle structures input.
- [#35] - Add module MUSCLE5_SUPER5.
- [#59] - Add support for passing structure template in samplesheet.
- [#77] - Add module zip.
- [#93] - Add multiqc basic support. Add custom params validation. Add basic Shiny app.
- [#100] - Add support for optional stats and evals. Clean tests.
- [#110] - Add Readme documentation. Add nf-test for the pipeline.
- [#76] - Add reading of trace files for Shiny app.
- [#99] - Add check for conflicting input parameters for stats and eval.
- [#117] - Add data for full datasets.
- [#90] - Add TCS evaluation metric.
- [#89] - Add collection of plddt metrics in stats subworkflow.
- [#96] - Add collection of number of gaps in eval subworkflow.
- [#84] - Update Metromap.
- [#139] - Add Foldmason.
- [#146] - Only show additional process tags when they exists and use the same ubuntu version in all modules.
- [#145] - Add consensus MSA.
- [#147] - Add small testing profile + some fixes of the Shiny app.
- [#148] - Add UPP module.
- [#150] - Update modules and readme for pre-release.
- [#61] - Fix filtering of unique guidetree compuation instructions: no double GT computation.
- [#57] - Update naming in the naming of the meta channels.
- [#66] - README: add new metromap and available tool list.
- [#54] - Update modules versions from nf-core tools.
- [#80] - Update modules versions from nf-core tools with nf-test.
- [#32] - Update Stats workflow with nf-core modules for merging.
- [#81] - Update Eval workflow with nf-core modules for merging.
- [#111] - Fix linting warnings (mostly versions).
- [#134] - Code revision for release preparation.
- [#138] - MultiQC as nf-core module and fix visualization.
- [#152] - Ignore kalign error 132 and print a warning (incompatibility with some CPU types).
- [#153] - Fix time parser in Shiny app.
- [#154] - Fix metromap.
- [#155] - Fix typos.