Merge pull request #322 from jonasscheid/ms2pip_cachedir

Add option to specify MS2PIP model directory
nf-core · May 28, 2024 · 2b22ec7 · 2b22ec7
2 parents 4a4f835 + 3dd15c0
commit 2b22ec7
Show file tree

Hide file tree

Showing 5 changed files with 15 additions and 2 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -8,6 +8,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### `Added`
 
 - Added MS²Rescore module with the underlying python CLI [#293](https://github.com/nf-core/mhcquant/pull/293)
+- Added new flag `ms2pip_model_dir`, which allows specifying a cache directory for ms2pip models [#322](https://github.com/nf-core/mhcquant/pull/322)
 
 ### `Fixed`
 

diff --git a/bin/ms2rescore_cli.py b/bin/ms2rescore_cli.py
@@ -37,6 +37,7 @@ def parse_cli_arguments_to_config(**kwargs):
                 config["ms2rescore"]["feature_generators"]["ms2pip"] = {
                     "model": kwargs["ms2pip_model"],
                     "ms2_tolerance": kwargs["ms2_tolerance"],
+                    "model_dir": kwargs["ms2pip_model_dir"],
                 }
             if "deeplc" in feature_generators:
                 config["ms2rescore"]["feature_generators"]["deeplc"] = {
@@ -144,6 +145,7 @@ def filter_out_artifact_psms(
     default="",
 )
 @click.option("-pipm", "--ms2pip_model", help="MS²PIP model (default: `Immuno-HCD`)", type=str, default="Immuno-HCD")
+@click.option("-pipmdir", "--ms2pip_model_dir", help="Path to directory, which holds pre-downloaded MS²PIP models", type=str, default=None)
 @click.option(
     "-ms2tol", "--ms2_tolerance", help="Fragment mass tolerance [Da](default: `0.02`)", type=float, default=0.02
 )

diff --git a/conf/modules.config b/conf/modules.config
@@ -215,6 +215,7 @@ process {
         ext.args   = [
             "--ms2_tolerance ${2 * params.fragment_mass_tolerance}",
             "--ms2pip_model ${params.ms2pip_model}",
+            "--ms2pip_model_dir ${params.ms2pip_model_dir}",
             "--rescoring_engine ${params.rescoring_engine}",
             params.feature_generators.trim() ? "--feature_generators ${params.feature_generators}" : ''
         ].join(' ').trim()

diff --git a/nextflow.config b/nextflow.config
@@ -50,6 +50,7 @@ params {
     rescoring_engine                = 'percolator'
     feature_generators              = 'deeplc,ms2pip'
     ms2pip_model                    = 'Immuno-HCD'
+    ms2pip_model_dir                = null
     deeplc_calibration_set_size     = 0.15
 
     // Percolator settings

diff --git a/nextflow_schema.json b/nextflow_schema.json
@@ -275,6 +275,7 @@
                     "enum": [
                         "Immuno-HCD",
                         "timsTOF",
+                        "timsTOF2023",
                         "CID",
                         "CIDch2",
                         "CID-TMT",
@@ -285,14 +286,21 @@
                         "iTRAQ",
                         "iTRAQphospho"
                     ],
-                    "description": "Specify the ms2pip model that should be used for rescoring. Checkout the ms2pip documentation for available models."
+                    "description": "Specify the MS²PIP model that should be used for rescoring. Checkout the MS²PIP documentation for available models."
+                },
+                "ms2pip_model_dir": {
+                    "type": "string",
+                    "format": "directory-path",
+                    "fa_icon": "fas fa-file-code",
+                    "hidden": true,
+                    "description": "Specify a local directory holding at least the MS²PIP models specified in `ms2pip_model`."
                 },
                 "deeplc_calibration_set_size": {
                     "type": "number",
                     "fa_icon": "fas fa-head-side-gear",
                     "hidden": true,
                     "default": 0.15,
-                    "description": "Specify the number or percentage of PSMs that should be used for calibration of the deeplc model."
+                    "description": "Specify the number or percentage of PSMs that should be used for calibration of the DeepLC model."
                 },
                 "fdr_level": {
                     "type": "string",