[New Version]: MS2DeepScore: 2.0.0 -> 2.5.3

tools/ms2deepscore/macros.xml

@@ -1,9 +1,14 @@
 <macros>
-    <token name="@TOOL_VERSION@">2.0.0</token>
-    <token name="@ONNX_VERSION@">1.16.2</token>
+    <token name="@TOOL_VERSION@">2.6.0</token>
+    <token name="@ONNX_VERSION@">1.19.1</token>
 
     <xml name="creator">
         <creator>
+            <person
+                givenName="Helge"
+                familyName="Hecht"
+                url="https://github.com/hechth"
+                identifier="0000-0001-6744-996X" />
             <person
                 givenName="Zargham"
                 familyName="Ahmad"
@@ -16,6 +21,13 @@
         </creator>
     </xml>
 
+    <xml name="environment_variables">
+        <environment_variables>
+            <environment_variable name="NUMBA_CACHE_DIR">tmp</environment_variable>
+            <environment_variable name="MPLCONFIGDIR">tmp</environment_variable>
+        </environment_variables>
+    </xml>
+
     <xml name="edam">
         <xrefs>
             <xref type="bio.tools">ms2deepscore</xref>
@@ -56,7 +68,7 @@
     <xml name="config_generator">
         <section name="model_structure" title="Model Structure" expanded="true">
             <repeat name="layers" title="Layer" min="1" default="1" >
-                <param name="dims" type="integer" label="Dimensions" min = "0" value="2000" help="Size of the in-between layer to add." />
+                <param name="dims" type="integer" label="Dimensions" min="1" value="2000" max="10000" help="Size of the in-between layer to add." />
             </repeat>
             <param name="embedding_dim" type="integer" label="Embedding Dimension" value="400" help="The dimension of the final embedding layer." />
             <param name="ionisation_mode" type="select" label="Ionisation Mode">
@@ -67,19 +79,28 @@
         </section>
 
         <section name="tensorization_settings" title="Tensorization Settings" expanded="true">
-            <param name="min_mz" type="integer" label="Min m/z" value="10" />
-            <param name="max_mz" type="integer" label="Max m/z" value="1000" />
-            <param name="mz_bin_width" type="float" label="m/z Bin Width" value="0.1" />
-            <param name="intensity_scaling" type="float" label="Intensity Scaling" value="0.5" />
-            <param name="fingerprint_type" type="text" value="daylight" label="Fingerprint Type" help="The fingerprint type that should be used for tanimoto score calculations." />
-            <param name="fingerprint_nbits" type="integer" label="Fingerprint Number of Bits" value="2048" help="The number of bits to use for the fingerprint." />
+            <param name="min_mz" type="integer" label="Min m/z" min="0" max="1000" value="10" 
+                help="Minimum m/z value to consider for tensorization." />
+            <param name="max_mz" type="integer" label="Max m/z" min="0" max="2000" value="1000" 
+                help="Maximum m/z value to consider for tensorization." />
+            <param name="mz_bin_width" type="float" label="m/z Bin Width" min="0" max="1" value="0.1" 
+                help="Width of each m/z bin for tensorization." />
+            <param name="intensity_scaling" type="float" label="Intensity Scaling" min="0" max="1" value="0.5" 
+                help="Scaling factor applied to intensities during tensorization." />
+            <param name="fingerprint_type" type="text" value="daylight" label="Fingerprint Type"
+                help="The fingerprint type that should be used for tanimoto score calculations." />
+            <param name="fingerprint_nbits" type="integer" label="Fingerprint Number of Bits" value="2048"
+                help="The number of bits to use for the fingerprint." />
         </section>
 
 
         <section name="training_settings" title="Training Settings" expanded="false">
-            <param name="dropout_rate" type="float" label="Dropout Rate" value="0.0" />
-            <param name="learning_rate" type="float" label="Learning Rate" value="0.00025" />
-            <param name="epochs" type="integer" label="Epochs" value="250" />
+            <param name="dropout_rate" type="float" label="Dropout Rate" value="0.0" 
+                help="Dropout rate to use during training to prevent overfitting." />
+            <param name="learning_rate" type="float" label="Learning Rate" value="0.00025" 
+                help="Learning rate for the optimizer during training." />
+            <param name="epochs" type="integer" label="Epochs" value="250" 
+                help="Number of training epochs." />
             <param name="patience" type="integer" label="Patience" value="20" help="How long the model should keep training if validation does not improve" />
             <param name="loss_function" type="select" label="Loss Function">
                 <option value="mse" selected="true">Mean Squared Error (mse)</option>
@@ -88,11 +109,23 @@
                 <option value="risk_mae">Risk Aware MAE (risk_aware_mae)</option>
                 <option value="risk_mse">Risk Aware MSE (risk_aware_mse)</option>
             </param>
-            <param name="weighting_factor" type="integer" label="Weighting Factor" value="0" />
+            <param name="weighting_factor" type="integer" label="Weighting Factor" value="0" 
+                help="Weighting factor for the loss function." />
             <param name="batch_size" type="integer" value="32" label="Batch Size" help="Number of pairs per batch" />
-            <param name="average_pairs_per_bin" type="integer" value="20" label="Average pairs per bin" help="The aimed average number of pairs of spectra per spectrum in each bin." />
+            <!-- <param name="average_pairs_per_bin" type="integer" value="20" label="Average pairs per bin" help="The aimed average number of pairs of spectra per spectrum in each bin." /> -->
             <param name="random_seed" type="text" label="Random seed" value="None" help="Specify random seed for reproducible random number generation." />
         </section>
+
+        <section name="data_generator_settings" title="Data Generator Settings" expanded="false">
+            <param name="average_inchikey_sampling_count" type="integer" label="Average InChIKey Sampling Count" value="100"
+                help="The average number of inchikeys that should be sampled for each bin. This is used to create pairs of spectra." />
+            <param name="max_inchikey_sampling_count" type="integer" label="Max InChIKey Sampling Count" value="110"
+                help="The maximum number of inchikeys that should be sampled for each bin. This is used to create pairs of spectra." />
+            <param name="max_pairs_per_bin" type="integer" label="Max Pairs per Bin" value="200"
+                help="The maximum number of pairs of spectra that should be created for each bin. This is used to create pairs of spectra." />
+            <param name="max_pair_resampling" type="integer" label="Max Pair Resampling" value="10000000"
+                help="The maximum number of times a pair of spectra can be resampled. This is used to create pairs of spectra." />
+        </section>
     </xml>
 
     <xml name="citations">

tools/ms2deepscore/ms2deepscore_config_generator.xml

@@ -1,4 +1,4 @@
-<tool id="ms2deepscore_config_generator" name="MS2DeepScore Configuration Generator" version="@TOOL_VERSION@+galaxy0">
+<tool id="ms2deepscore_config_generator" name="MS2DeepScore Configuration Generator" version="@TOOL_VERSION@+galaxy0" profile="23.0" license="MIT">
     <description>Generates model parameters for MS2DeepScore in JSON format</description>
     <macros>
         <import>macros.xml</import>
@@ -13,7 +13,10 @@
     <command detect_errors="exit_code"><![CDATA[
         python3 ${python_wrapper}
     ]]></command>
-    <configfiles>
+
+    <expand macro="environment_variables" />
+
+<configfiles>
     <configfile name="python_wrapper">
 import numpy as np
 from typing import Optional
@@ -38,11 +41,13 @@ params = {
     "mz_bin_width": $tensorization_settings.mz_bin_width,
     "intensity_scaling": $tensorization_settings.intensity_scaling,
     "batch_size": $training_settings.batch_size,
-    "average_pairs_per_bin": $training_settings.average_pairs_per_bin,
-    "same_prob_bins": np.array([(0, 0.2), (0.2, 1.0)]),
     "random_seed": random_seed,
     "fingerprint_type": "$tensorization_settings.fingerprint_type",
-    "fingerprint_nbits": $tensorization_settings.fingerprint_nbits
+    "fingerprint_nbits": $tensorization_settings.fingerprint_nbits,
+    "average_inchikey_sampling_count": $data_generator_settings.average_inchikey_sampling_count,
+    "max_inchikey_sampling": $data_generator_settings.max_inchikey_sampling_count,
+    "max_pairs_per_bin": $data_generator_settings.max_pairs_per_bin,
+    "max_pair_resampling": $data_generator_settings.max_pair_resampling,
 }
 
 settings = SettingsMS2Deepscore(**params)
@@ -55,19 +60,31 @@ settings.save_to_file("$output_file")
     </inputs>
 
     <outputs>
-        <data name="output_file" format="json" label="Model Parameter JSON" />
+        <data name="output_file" format="json" label="MS2DeepScore Model Configuration" />
     </outputs>
 
     <tests>
         <test expect_num_outputs="1">
-            <param name="layers_0|dims" value="20"/>
-            <param name="layers_1|dims" value="20"/>
-            <param name="embedding_dim" value="15" />
-            <param name="ionisation_mode" value="negative" />
-            <param name="epochs" value="2" />
-            <param name="batch_size" value="2" />
-            <param name="average_pairs_per_bin" value="2" />
-            <param name="random_seed" value="42"/>
+            <section name="model_structure">
+                <repeat name="layers">
+                    <param name="dims" value="20" />
+                </repeat>
+                <repeat name="layers">
+                    <param name="dims" value="20" />
+                </repeat>
+                <param name="embedding_dim" value="15" />
+                <param name="ionisation_mode" value="negative" />
+            </section>
+            <section name="training_settings">
+                <param name="epochs" value="2" />
+                <param name="batch_size" value="2" />
+                <param name="random_seed" value="42"/>
+                <param name="patience" value="2"/>
+            </section>
+            <section name="data_generator_settings">
+                <param name="average_inchikey_sampling_count" value="0"/>
+                <param name="max_pair_resampling" value="100"/>
+            </section>
             <output name="output_file" value="Model_Parameter_JSON.json" ftype="json" compare="diff" lines_diff="2"/>
         </test>
     </tests>

tools/ms2deepscore/ms2deepscore_similarity.xml

@@ -1,4 +1,4 @@
-<tool id="ms2deepscore_similarity" name="MS2DeepScore Similarity" version="@TOOL_VERSION@+galaxy0">
+<tool id="ms2deepscore_similarity" name="MS2DeepScore Similarity" version="@TOOL_VERSION@+galaxy0" profile="23.0" license="MIT">
     <description>Compute similarity scores using a pre-trained MS2DeepScore model</description>
     <macros>
         <import>macros.xml</import>
@@ -14,6 +14,9 @@
     <command detect_errors="exit_code"><![CDATA[
         python3 ${python_wrapper}
     ]]></command>
+
+    <expand macro="environment_variables"/>
+
 <configfiles>
 <configfile name="python_wrapper">
 @init_logger@
@@ -71,16 +74,20 @@ scores.to_json("$similarity_scores")
 
     <tests>
         <test expect_num_outputs="1">
-            <param name="use_scores" value="False"/>
-            <param name="references" value="inp_filtered_library.msp" ftype="msp"/>
-            <param name="queries" value="inp_filtered_spectra.msp" ftype="msp"/>
+            <conditional name="scores">
+                <param name="use_scores" value="False"/>
+                <param name="references" value="inp_filtered_library.msp" ftype="msp"/>
+                <param name="queries" value="inp_filtered_spectra.msp" ftype="msp"/>
+            </conditional>
             <param name="model" value="Trained_model.onnx" ftype="onnx"/>
             <param name="model_param" value="Model_Parameter_JSON.json" ftype="json"/>
             <output name="similarity_scores" value="msp_json_score_out.json" ftype="json" compare="sim_size" />
         </test>
         <test expect_num_outputs="1">
-            <param name="use_scores" value="True"/>
-            <param name="scores_in" value="ri_match_60.json" ftype="json"/>
+            <conditional name="scores">
+                <param name="use_scores" value="True"/>
+                <param name="scores_in" value="ri_match_60.json" ftype="json"/>
+            </conditional>
             <param name="model" value="Trained_model.onnx" ftype="onnx"/>
             <param name="model_param" value="Model_Parameter_JSON.json" ftype="json"/>
             <output name="similarity_scores" value="usescore_json_score_out.json" ftype="json" compare="sim_size" />

tools/ms2deepscore/ms2deepscore_training.xml

@@ -1,4 +1,4 @@
-<tool id="ms2deepscore_training" name="MS2DeepScore Model Training" version="@TOOL_VERSION@+galaxy0">
+<tool id="ms2deepscore_training" name="MS2DeepScore Model Training" version="@TOOL_VERSION@+galaxy0" profile="23.0" license="MIT">
     <description>Compute similarity scores using a pre-trained MS2DeepScore model</description>
     <macros>
         <import>macros.xml</import>
@@ -16,6 +16,9 @@
         cp $spectra processing/input."$spectra.ext";
         python3 ${python_wrapper}
     ]]></command>
+
+    <expand macro="environment_variables"/>
+
 <configfiles>
 <configfile name="python_wrapper">
 import onnx
@@ -63,15 +66,15 @@ torch.onnx.export(
     </inputs>
 
     <outputs>
-        <data label="Trained model" name="onnx_trained_model" format="onnx"/>
+        <data name="onnx_trained_model" format="onnx"/>
     </outputs>
 
     <tests>
         <test expect_num_outputs="1">
-            <param name="spectra" value="clean_spectra.mgf" ftype="mgf"/>
-            <param name="model_param" value="Model_Parameter_JSON.json" ftype="json" />
-            <param name="validation_split_fraction" value="5"/>
-            <output name="onnx_trained_model" value="Trained_model.onnx" ftype="onnx" compare="sim_size"/>
+            <param name="spectra" ftype="msp" location='https://galaxy-umsa.grid.cesnet.cz/api/datasets/fa57798f4c6835f5/display?to_ext=msp'/>
+            <param name="model_param" value="Model_Parameter_JSON_test.json" ftype="json" />
+            <param name="validation_split_fraction" value="20"/>
+            <output name="onnx_trained_model" location='https://galaxy-umsa.grid.cesnet.cz/api/datasets/feaa5809758067ab/display?to_ext=onnx' ftype="onnx" compare="sim_size"/>
         </test>
     </tests>
 

tools/ms2deepscore/test-data/Galaxy4-[Model Parameter JSON].json

@@ -0,0 +1,49 @@
+{
+    "base_dims": [
+        2000
+    ],
+    "embedding_dim": 400,
+    "ionisation_mode": "both",
+    "train_binning_layer": false,
+    "train_binning_layer_group_size": 20,
+    "train_binning_layer_output_per_group": 2,
+    "dropout_rate": 0.01,
+    "learning_rate": 0.00025,
+    "epochs": 10,
+    "patience": 5,
+    "loss_function": "mse",
+    "weighting_factor": 0,
+    "model_file_name": "ms2deepscore_model.pt",
+    "history_plot_file_name": "history.svg",
+    "time_stamp": "2025_06_12_17_53_28",
+    "min_mz": 10,
+    "max_mz": 1000,
+    "mz_bin_width": 0.1,
+    "intensity_scaling": 0.5,
+    "additional_metadata": [],
+    "batch_size": 64,
+    "num_turns": 1,
+    "shuffle": true,
+    "use_fixed_set": false,
+    "average_pairs_per_bin": 20,
+    "max_pairs_per_bin": 100,
+    "same_prob_bins": [
+        [
+            0.0,
+            0.2
+        ],
+        [
+            0.2,
+            1.0
+        ]
+    ],
+    "include_diagonal": true,
+    "random_seed": null,
+    "fingerprint_type": "daylight",
+    "fingerprint_nbits": 2048,
+    "augment_removal_max": 0.2,
+    "augment_removal_intensity": 0.2,
+    "augment_intensity": 0.2,
+    "augment_noise_max": 10,
+    "augment_noise_intensity": 0.02
+}

tools/ms2deepscore/test-data/Model_Parameter_JSON.json

@@ -5,18 +5,19 @@
     ],
     "embedding_dim": 15,
     "ionisation_mode": "negative",
+    "activation_function": "relu",
     "train_binning_layer": false,
     "train_binning_layer_group_size": 20,
     "train_binning_layer_output_per_group": 2,
     "dropout_rate": 0.0,
     "learning_rate": 0.00025,
     "epochs": 2,
-    "patience": 20,
+    "patience": 2,
     "loss_function": "mse",
     "weighting_factor": 0,
     "model_file_name": "ms2deepscore_model.pt",
     "history_plot_file_name": "history.svg",
-    "time_stamp": "2024_08_16_07_50_22",
+    "time_stamp": "2025_06_23_11_16_54",
     "min_mz": 10,
     "max_mz": 1000,
     "mz_bin_width": 0.1,
@@ -26,20 +27,55 @@
     "num_turns": 1,
     "shuffle": true,
     "use_fixed_set": false,
-    "average_pairs_per_bin": 2,
-    "max_pairs_per_bin": 100,
+    "average_inchikey_sampling_count": 0,
+    "max_inchikey_sampling": 110,
+    "max_pairs_per_bin": 200,
     "same_prob_bins": [
         [
-            0.0,
-            0.2
+            0.8,
+            0.9
         ],
         [
-            0.2,
+            0.7,
+            0.8
+        ],
+        [
+            0.9,
             1.0
+        ],
+        [
+            0.6,
+            0.7
+        ],
+        [
+            0.5,
+            0.6
+        ],
+        [
+            0.4,
+            0.5
+        ],
+        [
+            0.3,
+            0.4
+        ],
+        [
+            0.2,
+            0.3
+        ],
+        [
+            0.1,
+            0.2
+        ],
+        [
+            -0.01,
+            0.1
         ]
     ],
     "include_diagonal": true,
+    "val_spectra_per_inchikey": 1,
     "random_seed": 42,
+    "max_pair_resampling": 100,
     "fingerprint_type": "daylight",
     "fingerprint_nbits": 2048,
     "augment_removal_max": 0.2,