New TCR-Epitope Binding Affinity Prediction Task

tdc/metadata.py

@@ -144,6 +144,7 @@
 
 catalyst_dataset_names = ['uspto_catalyst']
 
+tcr_epi_dataset_names = ['weber']
 
 ####################################
 # generation
@@ -358,7 +359,8 @@
 								"DrugSyn",
 								"MTI",
 								"GDA",
-								"Catalyst"],
+								"Catalyst",
+								"TCR_Epitope_Binding"],
 				'generation': ["RetroSyn",
 								"Reaction",
 								"MolGen"
@@ -395,7 +397,8 @@ def get_task2category():
 				"Catalyst": catalyst_dataset_names, 
 				"CRISPROutcome": crisproutcome_dataset_names,
 				"test_single_pred": test_single_pred_dataset_names,
-				"test_multi_pred": test_multi_pred_dataset_names
+				"test_multi_pred": test_multi_pred_dataset_names,
+				"TCREpitopeBinding": tcr_epi_dataset_names
 				}
 
 benchmark_names = {"admet_group": admet_benchmark,
@@ -500,7 +503,8 @@ def get_task2category():
  'leenay':'tab',
  'test_single_pred': 'tab',
  'test_multi_pred': 'tab',
- 'gdsc_gene_symbols': 'tab'}
+ 'gdsc_gene_symbols': 'tab',
+ 'weber': 'tab'}
 
 name2id = {'bbb_adenot': 4259565,
  'bbb_martins': 4259566,
@@ -586,7 +590,8 @@ def get_task2category():
  'leenay':4279966,
  'test_single_pred': 4832455,
  'test_multi_pred': 4832456,
- 'gdsc_gene_symbols': 5255026}
+ 'gdsc_gene_symbols': 5255026,
+ 'weber': 5790963}
 
 oracle2type = {'drd2': 'pkl', 
 			   'jnk3': 'pkl', 

tdc/multi_pred/__init__.py

@@ -8,4 +8,5 @@
 from .mti import MTI
 from .peptidemhc import PeptideMHC
 from .ppi import PPI
-from .test_multi_pred import TestMultiPred
+from .test_multi_pred import TestMultiPred
+from .tcr_epi import TCREpitopeBinding

tdc/multi_pred/tcr_epi.py

@@ -0,0 +1,41 @@
+# -*- coding: utf-8 -*-
+# Author: TDC Team
+# License: MIT
+
+import warnings
+warnings.filterwarnings("ignore")
+import sys
+
+from ..utils import print_sys
+from ..utils.load import download_wrapper, pd_load
+from . import bi_pred_dataset, multi_pred_dataset
+from ..metadata import dataset_names
+
+class TCREpitopeBinding(multi_pred_dataset.DataLoader):
+
+    """Data loader class to load datasets in T cell receptor (TCR) Specificity Prediction Task. 
+    More info: 
+
+    Task Description: Given the TCR and epitope sequence, predict binding probability.
+
+    Args:
+        name (str): the dataset name.
+        path (str, optional): 
+            The path to save the data file, defaults to './data'
+        print_stats (bool, optional): 
+            Whether to print basic statistics of the dataset, defaults to False
+
+    """
+
+    def __init__(self, name, path='./data', print_stats=False):
+        """Create TCR Specificity Prediction dataloader object
+        """
+        super().__init__(name, path, print_stats,
+                         dataset_names=dataset_names["TCREpitopeBinding"])
+        self.entity1_name = 'TCR'
+        self.entity2_name = 'Epitope'
+
+        if print_stats:
+            self.print_stats()
+
+        print('Done!', flush=True, file=sys.stderr)