Add GREMLIN-DCA avGFP DirectEvo: ProSST to launch.json

Author: niklases

.gitignore

@@ -421,3 +421,4 @@ datasets/AVGFP/Pickles/*
 datasets/AVGFP/DCA_Hybrid_Model_Performance_ESM1v_no_ML.png
 datasets/AVGFP/DCA_Hybrid_Model_Performance_ProSST_no_ML.png
 datasets/AVGFP/HYBRIDgremlinesm_DE_trajectories.png
+datasets/AVGFP/HYBRIDgremlinprosst_DE_trajectories.png

.vscode/launch.json

@@ -369,6 +369,25 @@
             ]
         },
 
+        {
+            "name": "Python: PyPEF hybrid/only-PS-zero-shot GREMLIN-DCA avGFP DirectEvo: ProSST",
+            "type": "debugpy",
+            "request": "launch",
+            "env": {"PYTHONPATH": "${workspaceFolder}"},
+            "program": "${workspaceFolder}/pypef/main.py",
+            "console": "integratedTerminal",
+            "justMyCode": true,
+            "cwd": "${workspaceFolder}/datasets/AVGFP/",
+            "args": [
+                "hybrid", 
+                "directevo",
+                "-m", "HYBRIDgremlinprosst", 
+                "--wt", "P42212_F64L.fasta",
+                "--params", "GREMLIN",
+                "--temp", "0.1"
+            ]
+        },
+
         {   // PLMC zero-shot steps:
             // 1. $pypef param_inference --params uref100_avgfp_jhmmer_119_plmc_42.6.params
             // 2. $pypef hybrid -t TS.fasl --params PLMC

pypef/utils/directed_evolution.py

@@ -240,13 +240,14 @@ def in_silico_de(self):
                 if wt_prediction is None or wt_prediction == 'skip':
                     wt_prediction = 'skip'
                     while wt_prediction == 'skip':
-                        rand_pos = random.randint(0, len(self.s_wt))
+                        rand_pos = random.randint(0, len(self.s_wt) - 1)
                         wt_mut = self.s_wt[rand_pos] + str(rand_pos) + self.s_wt[rand_pos]
+                        print(f"Trying to get WT fitness: {wt_mut}...")
                         wt_prediction = predict(  # AAidx, OneHot, or DCA-based pure ML prediction
                             path=self.path,
                             model=self.model,
                             encoding=self.encoding,
-                            variants=np.atleast_1d(wt_mut),
+                            variants=np.atleast_1d(wt_mut),   # WT, e.g. F17F
                             sequences=np.atleast_1d(self.s_wt),
                             no_fft=self.no_fft,
                             couplings_file=self.dca_encoder
@@ -272,14 +273,14 @@ def in_silico_de(self):
                 if wt_prediction is None or wt_prediction == 'skip':
                     wt_prediction = 'skip'
                     while wt_prediction == 'skip':
-                        rand_pos = random.randint(0, len(self.s_wt))
+                        rand_pos = random.randint(0, len(self.s_wt) - 1)
                         wt_mut = self.s_wt[rand_pos] + str(rand_pos) + self.s_wt[rand_pos]
+                        print(f"Trying to get WT fitness: {wt_mut}...")
                         wt_prediction = predict_directed_evolution(
                             encoder=self.dca_encoder,
-                            variant=self.s_wt[int(new_variant[:-1]) - 1] + new_variant[:-1] + 
-                                self.s_wt[int(new_variant[:-1]) - 1],  # WT, e.g. F17F
+                            variant=wt_mut,  # WT, e.g. F17F
                             variant_sequence=self.s_wt,
-                            hybrid_model_data_pkl=self.model
+                            hybrid_model_data_pkl=self.model  # TODO: Add global model
                         )
                 if self.de_step_counter == 0:
                     logger.info(