New testing scheme

clients/python/test/Prosit/test_Prosit_2024_intensity_XL_NMS2.py

@@ -0,0 +1,23 @@
+from test.server_config import SERVER_GRPC, SERVER_HTTP
+from pathlib import Path
+from test.lib import (
+    lib_test_available_grpc,
+    lib_test_available_http,
+    lib_test_inference,
+)
+
+
+# To ensure MODEL_NAME == test_<filename>.py
+MODEL_NAME = Path(__file__).stem.replace("test_", "")
+
+
+def test_available_http():
+    lib_test_available_http(MODEL_NAME, SERVER_HTTP)
+
+
+def test_available_grpc():
+    lib_test_available_grpc(MODEL_NAME, SERVER_GRPC)
+
+
+def test_inference():
+    lib_test_inference(MODEL_NAME, SERVER_GRPC)

clients/python/test/Prosit/test_Prosit_2024_intensity_XL_NMS2_core.py

@@ -0,0 +1,23 @@
+from test.server_config import SERVER_GRPC, SERVER_HTTP
+from pathlib import Path
+from test.lib import (
+    lib_test_available_grpc,
+    lib_test_available_http,
+    lib_test_inference,
+)
+
+
+# To ensure MODEL_NAME == test_<filename>.py
+MODEL_NAME = Path(__file__).stem.replace("test_", "")
+
+
+def test_available_http():
+    lib_test_available_http(MODEL_NAME, SERVER_HTTP)
+
+
+def test_available_grpc():
+    lib_test_available_grpc(MODEL_NAME, SERVER_GRPC)
+
+
+def test_inference():
+    lib_test_inference(MODEL_NAME, SERVER_GRPC)

clients/python/test/lib.py

@@ -15,7 +15,7 @@ def lib_test_available_grpc(model_name, server_grpc):
     assert client._is_model_ready() is None
 
 
-def lib_test_inference(model_name, server_grpc, atol=1e-6):
+def lib_test_inference(model_name, server_grpc, atol=1e-4):
     files = glob(f"**/arr-{model_name}-*.npy", recursive=True)
     data = {Path(f).stem.split("-")[-1]: np.load(f) for f in files}
 

models/Prosit/Prosit_2024_intensity_XL_NMS2/config.pbtxt

@@ -0,0 +1,158 @@
+max_batch_size: 1000
+platform: "ensemble"
+input [
+  {
+   name: 'peptide_sequences_1',
+   data_type: TYPE_STRING,
+   dims: [-1]
+  },
+  {
+   name: 'peptide_sequences_2',
+   data_type: TYPE_STRING,
+   dims: [-1]
+  },
+  {
+    name: 'precursor_charges',
+    data_type: TYPE_INT32,
+    dims: [1],
+  },
+  {
+    name: 'collision_energies',
+    data_type: TYPE_FP32,
+    dims: [1],
+  }
+]
+output [
+  {
+   name: 'intensities',
+   data_type: TYPE_FP32,
+   dims: [174]
+  },
+ {
+   name: 'mz',
+   data_type: TYPE_FP32,
+   dims: [174]
+ },
+ {
+   name: 'annotation',
+   data_type: TYPE_STRING,
+   dims: [174]
+ }
+]
+
+ensemble_scheduling {
+  step [
+     {
+      model_name: "Prosit_Preprocess_charge"
+      model_version: 1
+      input_map {
+        key: "precursor_charges"
+        value: "precursor_charges"
+      },
+      output_map {
+        key: "precursor_charges_in:0"
+        value: "precursor_charges_in_preprocessed:0"
+      }
+    },
+    {
+      model_name: "XL_Prosit_Preprocess_peptide_NMS2"
+      model_version: 1
+      input_map {
+        key: "peptide_sequences_1"
+        value: "peptide_sequences_1"
+      },
+      input_map {
+        key: "peptide_sequences_2"
+        value: "peptide_sequences_2"
+      }
+      output_map {
+        key: "peptides_in_1:0"
+        value: "peptides_in_1:0"
+      },
+      output_map {
+        key: "peptides_in_2:0"
+        value: "peptides_in_2:0"
+      }
+    },
+    {
+      model_name: "Prosit_Preprocess_collision_energy"
+      model_version: 1
+      input_map {
+        key: "raw_collision_energy"
+        value: "collision_energies"
+      },
+      output_map {
+        key: "norm_collision_energy"
+        value: "norm_collision_energy"
+      }
+    },
+    {
+      model_name: "Prosit_2024_intensity_XL_NMS2_core"
+      model_version: 1
+      input_map {
+        key: "peptides_in1"
+        value: "peptides_in_1:0"
+      },
+      input_map {
+        key: "peptides_in2"
+        value: "peptides_in_2:0"
+      },
+      input_map {
+        key: "collision_energy_in"
+        value: "norm_collision_energy"
+      },
+      input_map {
+        key: "precursor_charge_in"
+        value: "precursor_charges_in_preprocessed:0"
+      }
+      output_map {
+        key: "out_1"
+        value: "out/Reshape:0"
+      }
+    },
+    {
+      model_name: "XL_Prosit_NMS2_2024_intensity_postprocess"
+      model_version: 1
+      input_map {
+        key: "peptides_in_1:0"
+        value: "peptide_sequences_1"
+      },
+      input_map {
+        key: "peptides_in_2:0"
+        value: "peptide_sequences_2"
+      },
+      input_map{
+        key: "precursor_charges_in:0"
+        value: "precursor_charges_in_preprocessed:0"
+      }
+      input_map{
+        key: "peaks_in:0",
+        value: "out/Reshape:0"
+      }
+      output_map {
+        key: "intensities"
+        value: "intensities"
+      }
+      output_map {
+        key: "mz"
+        value: "mz"
+      }
+    },
+    {
+      model_name: "XL_Prosit_NMS2_Helper_annotation"
+      model_version: 1
+      input_map {
+        key: "precursor_charges"
+        value: "precursor_charges"
+      },
+      input_map {
+        key: "peptide_sequences_1"
+        value: "peptide_sequences_1"
+      }
+      output_map {
+        key: "annotation"
+        value: "annotation"
+      }
+    }
+  ]
+}

models/Prosit/Prosit_2024_intensity_XL_NMS2/notes.yaml

@@ -0,0 +1,37 @@
+description: |
+  The HCD Prosit 2020 model was fine-tuned using using 251,449 sets of cross-linked peptides, specifically those created from DSSO and DSBU. 
+  The model's structure was adjusted to accommodate two input peptides and generate both non-xl and xl fragments in the output. 
+  The dataset was divided into three parts: training (80% - 198,294 sets), validation (10% - 26,610 sets), and test (10% - 26,545 sets).
+  This model achieved a SA (a measure of accuracy) of at least 0.83, which is a significant improvement compared to the HCD Prosit 2020 model's performance on cross-linked peptides, where the accuracy was SA ≥ 0.58
+
+
+tag: "Intensity"
+tag_url: "https://koina.wilhelmlab.org/"
+examples:
+  inputs:
+    [
+      {
+          "name": "peptide_sequences_1",
+          "httpdtype": "BYTES",
+          "shape": "[2,1]",
+          "data": '["DIADAVTAAGVEVAK[UNIMOD:1896]SEVR", "AGDQIQSGVDAAIK[UNIMOD:1896]PGNTLPMR"]'
+      },
+      {
+          "name": "peptide_sequences_2",
+          "httpdtype": "BYTES",
+          "shape": "[2,1]",
+          "data": '["NFLVPQGK[UNIMOD:1896]AVPATK", "SANIALVLYK[UNIMOD:1896]DGER"]'
+      },
+      {
+          "name": "precursor_charges",
+          "httpdtype": "INT32",
+          "shape": "[2,1]",
+          "data": '[3,4]'
+      },
+      {
+          "name": "collision_energies",
+          "httpdtype": "FP32",
+          "shape": "[2,1]",
+          "data": '[28, 28]'
+      },
+    ]

models/Prosit/Prosit_2024_intensity_XL_NMS2_core/1/.zenodo

@@ -0,0 +1,2 @@
+https://zenodo.org/records/11259344/files/model.savedmodel.zip?download=1
+md5:5e65b1e60013d94c1e957d62ad601b03

models/Prosit/Prosit_2024_intensity_XL_NMS2_core/config.pbtxt

@@ -0,0 +1 @@
+max_batch_size: 1000