Release v5.5.4

Release aiming at fixing two main issues:

1. In the case of having duplicated isoform and reference IDs rescue could crash (some tools like Bamboo or Flair assing the name of the refence to the isoforms in the transcriptome if a perfect match is found)
2. In the parallelization, the isoform_hits option was bugged, as well as a small bug with the corruption of a header in the abundance calculation.

What's Changed

• Add disk space cleanup to DockerHub release workflow by @Copilot in #549
• parallel issue fixed by @pabloati in #557
• Fixed parallelization isoforms_hits by @pabloati in #560
• Optimize GitHub Actions workflows with path filters and caching by @Copilot in #558

Full Changelog: v5.5.3...v5.5.4

Release v5.5.3

Quick release to fix some important issues that arose in SQANTI3 in the last days. The next release (hopelfuly) will implement major changes in the pipeline!

What's Changed

• Bugfix/rescue duplications 545 by @pabloati in #546
• Fix to saturation plots failing when isoforms lacked counts by @pabloati in #548
• Fix to the abundance count pipeline in the parallel mode, pointed out by #556 by @pabloati

Full Changelog: v5.5.2...v5.5.3

Release v5.5.2

Bugfix release aimed at fixing minor error in the filter report, compilation of conda packages and testing and the classification file

What's Changed

• Changed ujc_counts.csv file reads grouping by @CarlosBlancoGo in #518
• Fix docker chmod copyup by @TianYuan-Liu in #521
• tusco: remove species GTF dependency; use TSV + refGTF for plotting by @TianYuan-Liu in #520
• Changed ujc_counts.csv file reads grouping by @CarlosBlancoGo in #519
• Minor bug in the report by @alexpan00 in #525
• docker: fix build by removing manual compilation and fixing conda TOS by @TianYuan-Liu in #528
• TUSCO: Improve TSV parsing and update human gene list by @TianYuan-Liu in #529
• Fix conda environment Python 3.11.13 compatibility and add macOS CI testing by @TianYuan-Liu in #530
• made plotting step optional in sqanti reads by @fairliereese in #533
• Filter report and reference length update by @pabloati in #543
• Releases/v5.5.2 by @pabloati in #544

New Contributors

• @fairliereese made their first contribution in #533

Full Changelog: v5.5.1...v5.5.2

Release v5.5.1

Minor bugfixes from previous version plus a conda environment upgrade

What's Changed

• Fix: Update to python's version to 3.11.13, to prevent freezes in TD2, as found by @FabianJetzinger
• Fix: preprocess corrected.gtf with gffread before gtftools in make_UJ… by @pabloati in #494
• Fix check of categorical rules with multiple values by @FabianJetzinger in #497
• Bugfix/negative strand positions 493 by @pabloati in #506
• Bugfix/isoannot 500 by @pabloati in #510
• Releases/v5.5.1 by @pabloati in #511

Full Changelog: v5.5...v5.5.1

Release v5.5

🛠️ Minor Release

This minor release main change is the subsitution of GeneMarkS-T for TransDecoder2 to predict ORFs. This change was done due to the outdating of GMST and its constant issues that we could not solve. Moreover, some essential bugfixes and QOL improvements have been added to the repository. Read below for more information.

---

✨ Improvements

• Automatic detection of @SQ headers in SAM files.
• Added multiple checks for gene_name attributes in reference annotations when required.
• Improved error handling in rescue rules when no valid rules are provided.
• Significantly faster rescue file writing (no shell call overhead).
• Enhanced output management and reduced type errors in filterML.

---

🐞 Bugfixes

• Fixed logging: missing files now raise appropriate warnings or errors.
• Boolean values in the config file are now correctly detected by the wrapper.
• IsoAnnotLite now checks for the gene_name attribute in the reference.
• Fixed rescue output generation when no GTF is provided.
• Custom TP/TN sets and remove_columns in filterML are now handled correctly.
• SQANTI-reads adapted to the updated argument parser.
• Fixed string conversion issue in myQueryTranscripts.__str__() when printing None.
• TSS ratio counts have been corrected.
• CDS prediction bugs fixed.
• Fixed IO handling and logging behaviors.

---

🔧 Changes

• GMST has been replaced by TD2.
• SQANTI-reads updated to use the new argument format.
• Output filenames for CDS predictions have been made more descriptive.
• Examples updated for consistency with the new version.
• Rescue script reverted to its previous argument style.

---

📌 What's Changed (from PRs)

• Fix failure of myQueryTranscripts.__str__() when printing value of an invalid object [#455] by @diekhans
• TSS ratio counts fixed [#457](#457) by @pabloati
• IO handling and logging improved [#462] by @pabloati
• Rescue argparse fixes [#463] by @pabloati
• Fixed custom TP/TN sets and remove_columns [#465]by @FabianJetzinger
• Included SAM header detection [#468] by @pabloati
• Added user input validation on ´IsoAnnotLite´ and ´gene_name´ [#469] by @pabloati
• CDS prediction fixes [#470] by @pabloati
• Improved output management [#477] by @pabloati
• Type error fixes in filters [#478] by @pabloati
• ORF prediction enhancements [#486] by @pabloati
• Updated filter testing suite [#487] by @pabloati
• Version bump to v5.5 [#490] by @pabloati
• Small update to SQANTI-reads to standarize the format of the GTFs [#494] by @TianYuan-Liu
• Fix to SQANTI filter rules, to apply multioptions in categies as OR, rather than AND [#497] by @FabianJetzinger

---

🙌 New Contributor

• @FabianJetzinger made their first contribution in [#465](#465) 🎉

---

🔗 Full Changelog: v5.4...v5.5

Release 5.4

Introduction

This release introducessignificant changes in SQANTI3.

1. Introduciton of a loggin system --> Now, all theouptut to the console is controlled via a python logger. The external tools log is redirected to their respective files in the directory logs of each run.
2. Creation of a wrapper and a config file to run SQANTI3. This wrapper now works via a config file, thus some arguments and the way the parser works have been changed to be adapted to the wrapper
3. Reestructurarion of the filter module. Most of its scripts have been pythonized and the internal organization has been changed, with modifications to the file namings
4. Reestructuration of rescue module. Similar to filter, with most of the scirpts changed from R to python, eliminating many calls to external tools or the terminal
5. Fixed many bugs that were found along the way. See commits and pull requests for more info

What's Changed

• Adition of the wrapper to SQANTI by @pabloati in #392
• Junctions classification issue fixed by @pabloati in #398
• Bugfix/gtf order by @pabloati in #399
• Logger by @pabloati in #400
• Tests/junctions processing by @pabloati in #401
• Github actions: test docker build each commit, and publish to dockerhub on release by @Fabian-RY in #418
• Conda testing by @Fabian-RY in #419
• FIX: conda change in latest docker failed by @Fabian-RY in #420
• Switch attribute parser to regex by @SwiftSeal in #423
• Size one cake peak fix by @pabloati in #430
• Bugfix/sqanti reads by @pabloati in #432
• Bugfix/sqanti reads paula by @pabloati in #435
• Filter restructuration by @pabloati in #436
• docs: update README.md by @eltociear in #438
• Bugfix/gff isoannot by @pabloati in #439
• Rescue reorganization by @pabloati in #449
• Qc report by @pabloati in #450
• Fixed issue 444 by @pabloati in #451
• Logger filesystem by @pabloati in #453
• Releases/v5.4 by @pabloati in #454

New Contributors

• @SwiftSeal made their first contribution in #423
• @eltociear made their first contribution in #438

Full Changelog: v5.3.6...v5.4

v5.3.6

What's Changed

• Update sqanti3_filter.py by @sbresnahan in #397
• Updated the fl count parser to accept floats in single samples by @pabloati in #406
• Input file processing upgraded to accept any kind og GTF by @pabloati in #407
• Fixed report to accept classification with junction types missing by @pabloati in #408
• Issue 390, pallete and flair input fixed by @pabloati in #412
• GMST directory saved after parallelization by @pabloati in #414
• Fixed bug where the isoform hits file was not being propperly created and the flag killed the process, as in issue 431
• Release 5.3.6 update by @pabloati in #415

New Contributors

• @sbresnahan made their first contribution in #397

Full Changelog: v5.3.5...v5.3.6

v5.3.5

What's Changed

• Reduced the strictness for the arguments in SQANTI QC
• Fixed temporary files removal in parallelization as in issue 388
• Fixed subclassification of NIC between intergenic and genic intron as in issue 389
• update to reach master and make changes by @carolinamonzo in #393
• Sqanti reads devel updating utilities paths by @carolinamonzo in #394
• Updated example and tests to match the new improvements

Full Changelog: v5.3.4...v5.3.5

Release 5.3.4: Classification and subclassfication bug fixes

Through there release, the main issues fixed are related to the issues 381 and 379. As new features, SQANTI3 directly takes one bam file as SR_bam input, and the input format for the isoforms ID has been expanded to accept non-IsoSeq nomenclature.

What's Changed

• SR_bam now parses bam files directly by @pabloati in #383
• Updated the isoform sorting to accept a wider nomenclature by @pabloati in #384
• Fixed subclassification bug for NNC by @pabloati in #385
• changed the input gtf to output introns to the corrected cds gff by @CarlosBlancoGo in #370
• Sqanti reads devel merge compatibility fix from Carlos by @carolinamonzo in #386

New Contributors

• @CarlosBlancoGo made their first contribution in #370

Full Changelog: v5.3.3...v5.3.4

Release 5.3.3: GMST and dependencies bugs fixing

This release fixes an error where the HMM from GeneMarkS-T was saved in the script where SQANTI3 was being called. This caused issues when the parallelization was run, since the HMMs from different chunks tried to overwrite each other. Some minor changes were made to fix deprecated dependencies and versioning

What's Changed

• Fixed GMST so the HMM is saved in its directory by @pabloati in #380
• Updated deprecated dependencies (shutil to search for executables)
• Unified the versioning system

Full Changelog: v5.3.2...v5.3.3