Add support for using GFF instead of GTF

CHANGELOG.md

@@ -5,25 +5,27 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## v4.1.0dev
 
+- Add the `--gtf` parameter to allow the user to specify a GFF file as a reference (instead of a GTF file) ([#451](https://github.com/nf-core/scrnaseq/pull/451))
+
 ## v4.0.0 - 2025-03-10
 
-- Move `txp2gene` to `reference_genome_options` in schema as it is required by `kb_python` and `alevin` ([434](https://github.com/nf-core/scrnaseq/pull/434))
-- Fix of additional path splitting for `txp2gene` ([433](https://github.com/nf-core/scrnaseq/pull/433))
+- Move `txp2gene` to `reference_genome_options` in schema as it is required by `kb_python` and `alevin` ([#434](https://github.com/nf-core/scrnaseq/pull/434))
+- Fix of additional path splitting for `txp2gene` ([#433](https://github.com/nf-core/scrnaseq/pull/433))
 - Add a checker so that `--fb_reference` does not break the pipeline in case `ab` files are not used in `cellranger multi` sub-workflow.
-- Fix concatenation of multiple samples into the combined output AnnData ([416](https://github.com/nf-core/scrnaseq/pull/416))
-- Make sure STARsolo velocity output is added to the combined output AnnData, if `star_feature = 'Gene Velocyto'` ([417](https://github.com/nf-core/scrnaseq/pull/417))
+- Fix concatenation of multiple samples into the combined output AnnData ([#416](https://github.com/nf-core/scrnaseq/pull/416))
+- Make sure STARsolo velocity output is added to the combined output AnnData, if `star_feature = 'Gene Velocyto'` ([#417](https://github.com/nf-core/scrnaseq/pull/417))
 - Update cellbender module to latest nf-core version ([#419](https://github.com/nf-core/scrnaseq/pull/419/))
 - Add profile for gpu processes ([#419](https://github.com/nf-core/scrnaseq/pull/419/))
 - Update example usage command in README with valid reference genome parameter ([#339](https://github.com/nf-core/scrnaseq/issues/339))
 - Removed `--kb_filter` parameter. Kallisto filtering is triggered by default and can be turned off with `ext.args` ([#421](https://github.com/nf-core/scrnaseq/issues/421))
 - Document better that `cellbender` is used for empty drops calling and not the `emptydrops` method (([#420](https://github.com/nf-core/scrnaseq/issues/420)))
 - Add `--limitBAMsortRAM` to STARsolo alignment, to make sure BAM sorting memory scales with the task memory ([#430](https://github.com/nf-core/scrnaseq/pull/430))
-- Replace local modules for simpleaf, `SIMPLEAF_INDEX` and `SIMPLEAF_QUANT`, with their central modules from nf-core/modules, and update simpleaf subworkflows accordingly. ([424](https://github.com/nf-core/scrnaseq/pull/424))
-- Fix of additional path splitting for `txp2gene` ([433](https://github.com/nf-core/scrnaseq/pull/433))
-- Update documents related to `simpleaf`, `alevin`, `salmon`, and `alevin-fry` for consistency.([424](https://github.com/nf-core/scrnaseq/pull/424))
-- Rename the default aligner from `alevin` to `simpleaf` for consistency.([424](https://github.com/nf-core/scrnaseq/pull/424))
-- Update the `mtx_to_h5ad` template for `simpleaf` to start from the h5ad file generated by simpleaf.([424](https://github.com/nf-core/scrnaseq/pull/424))
-- Upgrade alevinqc from 1.12.1 to 1.18.0 to match the latest output file structure of simpleaf.([424](https://github.com/nf-core/scrnaseq/pull/424))
+- Replace local modules for simpleaf, `SIMPLEAF_INDEX` and `SIMPLEAF_QUANT`, with their central modules from nf-core/modules, and update simpleaf subworkflows accordingly. ([#424](https://github.com/nf-core/scrnaseq/pull/424))
+- Fix of additional path splitting for `txp2gene` ([#433](https://github.com/nf-core/scrnaseq/pull/433))
+- Update documents related to `simpleaf`, `alevin`, `salmon`, and `alevin-fry` for consistency.([#424](https://github.com/nf-core/scrnaseq/pull/424))
+- Rename the default aligner from `alevin` to `simpleaf` for consistency.([#424](https://github.com/nf-core/scrnaseq/pull/424))
+- Update the `mtx_to_h5ad` template for `simpleaf` to start from the h5ad file generated by simpleaf.([#424](https://github.com/nf-core/scrnaseq/pull/424))
+- Upgrade alevinqc from 1.12.1 to 1.18.0 to match the latest output file structure of simpleaf.([#424](https://github.com/nf-core/scrnaseq/pull/424))
 
 ## v3.0.0 - 2024-12-09
 

conf/modules.config

@@ -31,6 +31,16 @@ process {
         ]
     }
 
+    withName: 'GFFREAD' {
+        ext.args   = '--keep-exon-attrs -F -T'
+        publishDir = [
+            path: { "${params.outdir}/reference_genome" },
+            mode: params.publish_dir_mode,
+            enabled : params.save_reference,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+        ]
+    }
+
     if (!params.skip_cellbender) {
         withName: 'CELLBENDER_REMOVEBACKGROUND' {
             publishDir = [

docs/output.md

@@ -142,6 +142,8 @@ The pipeline also possess a subworkflow imported from scdownstream to perform fi
 - `extract_transcriptome`
   - When supplied with a `--fasta` genome fasta, this contains the extracted transcriptome
   - The GTF file supplied with `--gtf` is used to extract the transcriptome positions appropriately
+- `*.gtf`
+  - If only a `.gff` file was provided, the `.gtf` file generated by converting the `.gff` file will be saved here if using `--save-reference`.
 
 **Output directory: `results/${params.aligner}/mtx_conversions`**
 

docs/usage.md

@@ -146,7 +146,7 @@ you have to create a new cellranger-arc index ([see here](https://support.10xgen
 more information)
 
 If you decide to create a cellranger-arc index, then you need to create a config file to generate the index. The pipeline
-can do this autmatically for you if you provide a `--fasta`, `--gtf`, and an optional `--motif` file. However, you can
+can do this autmatically for you if you provide a `--fasta`, `--gtf` or `--gff`, and an optional `--motif` file. However, you can
 also decide to provide your own config file with `--cellrangerarc_config`, then you also have to specify with `--cellrangerarc_reference`
 the reference genome name that you have used and stated as _genome:_ in your config file.
 
@@ -279,7 +279,7 @@ The `sample` column must match the corresponding entry in the main samplesheet.
 #### Additional reference data
 
 - Cellranger multi needs a reference for **GEX and VDJ analysis**. They are calculated on the fly given the reference
-  files (`--fasta` and `--gtf`) provided, but users can also provide their own with: `--cellranger_index`
+  files (`--fasta`, and `--gtf` or `--gff`) provided, but users can also provide their own with: `--cellranger_index`
   and `--cellranger_vdj_index`, for GEX and VDJ, respectively.
 
   > When running cellranger multi, without any VDJ data, users can also skip VDJ automated ref building with: `--skip_cellrangermulti_vdjref`.

modules.json

@@ -62,7 +62,7 @@
                     },
                     "gffread": {
                         "branch": "master",
-                        "git_sha": "b1b959609bda44341120aed1766329909f54b8d0",
+                        "git_sha": "81880787133db07d9b4c1febd152c090eb8325dc",
                         "installed_by": ["modules"]
                     },
                     "gunzip": {