fixing test issues

src/api/file_atac_h5ad.yaml

@@ -1,5 +1,5 @@
 type: file
-example: resources/inference_datasets/op_atac.h5ad
+example: resources_test/inference_datasets/op_atac.h5ad
 info:
   label: multiomics atac
   summary: "Peak data for multiomics data."

src/api/file_evaluation_h5ad.yaml

@@ -1,5 +1,5 @@
 type: file
-example: resources/evaluation_datasets/op_perturbation.h5ad
+example: resources_test/evaluation_datasets/op_perturbation.h5ad
 info:
   label: perturbation
   summary: "Perturbation dataset for benchmarking."

src/api/file_rna_h5ad.yaml

@@ -1,5 +1,5 @@
 type: file
-example: resources/inference_datasets/op_rna.h5ad
+example: resources_test/inference_datasets/op_rna.h5ad
 info:
   label: multiomics rna
   summary: "RNA expression for multiomics data."

src/api/file_score.yaml

@@ -1,5 +1,5 @@
 type: file
-example: resources/scores/score.h5ad
+example: resources_test/scores/score.h5ad
 info:
   label: Score
   summary: "File indicating the score of a metric."

src/helper.py

@@ -86,14 +86,14 @@ def run_grn_inference():
     if True:
         par = {
             # 'methods': ["positive_control", "negative_control", "pearson_corr", "portia", "grnboost2", "ppcor", "scenic"],
-            'methods': ["positive_control", "negative_control", "pearson_corr"],
+            'methods': ["portia", "grnboost2"],
 
             'models_dir': 'resources/grn_models/replogle2/',
             'rna': 'resources/inference_datasets/replogle2.h5ad', 
             'rna_positive_control': 'resources/datasets_raw/replogle2.h5ad', 
             'num_workers': 20,
             'mem': "64GB",
-            'time': "48:00:00",
+            'time': "12:00:00",
             'normalize': False,
             # 'max_n_links': 10000,
         }
@@ -168,8 +168,8 @@ def run_grn_inference():
             full_tag += ["--partition=gpu", "--gres=gpu:1"]
 
         try:
-            # result = subprocess.run(['sbatch'] + full_tag + ['scripts/sbatch/grn_inference.sh', command], check=True, capture_output=True, text=True)
-            result = subprocess.run(['bash'] + ['scripts/sbatch/grn_inference.sh', command], check=True, capture_output=True, text=True)
+            result = subprocess.run(['sbatch'] + full_tag + ['scripts/sbatch/grn_inference.sh', command], check=True, capture_output=True, text=True)
+            # result = subprocess.run(['bash'] + ['scripts/sbatch/grn_inference.sh', command], check=True, capture_output=True, text=True)
 
             print(f"Job {method} submitted successfully.")
             print(result.stdout)  # Print the standard output

src/metrics/regression_1/main.py

@@ -115,7 +115,7 @@ def cross_validation(net, prturb_adata, par:dict):
         regr = lightgbm_wrapper(params, max_workers=par['num_workers'])
     else:
         raise ValueError(f"{reg_type} is not defined.")  
-
+    reg_models = []
     for group in verbose_tqdm(unique_groups, "Processing groups", 2, par['verbose']):
         mask_va = groups == group
         mask_tr = ~mask_va
@@ -125,7 +125,9 @@ def cross_validation(net, prturb_adata, par:dict):
             continue 
         regr.fit(X_tr, Y_tr)
         y_pred[mask_va & mask_shared_genes, :] = regr.predict(X[mask_va & mask_shared_genes, :])
-    return y_true, y_pred
+
+        reg_models.append(regr)
+    return y_true, y_pred, reg_models
 
 def regression_1(
             net: pd.DataFrame, 
@@ -162,9 +164,10 @@ def regression_1(
             prturb_adata_sub = prturb_adata[prturb_adata.obs.donor_id==donor_id,:]
         else:
             prturb_adata_sub = prturb_adata
-        y_true_sub, y_pred_sub = cross_validation(net_sub, prturb_adata_sub, par)
+        y_true_sub, y_pred_sub, _ = cross_validation(net_sub, prturb_adata_sub, par)
 
         score = r2_score(y_true_sub, y_pred_sub, multioutput='variance_weighted')
+
         score_list.append(score)
     mean_score_r2 = np.mean(score_list)
     output = dict(mean_score_r2=mean_score_r2)

src/metrics/script_all.py

@@ -5,19 +5,16 @@
 import os 
 
 ## VIASH START
-par = {
-  'reg_type': 'GB',
-  'models_dir': "resources/grn_models/replogle2",
-  'scores_dir': "output/",
+if True: #op 
+  par = {
+  'reg_type': 'ridge',
+  'models_dir': "resources/grn_models/op",
+  'scores_dir': "output/temp",
 
-  # 'models': [ 'collectri', 'negative_control', 'positive_control', 'pearson_corr', 'portia', 'ppcor', 'grnboost2', 'scenic', 'granie', 'scglue', 'celloracle', 'figr', 'scenicplus'],
-  'models': [ 'negative_control', 'positive_control', 'pearson_corr', 'portia', 'grnboost2', 'ppcor', 'scenic'],
-  # 'models': ['negative_control', 'positive_control', 'pearson_corr'],
+  'models': [ 'collectri', 'negative_control', 'positive_control', 'pearson_corr', 'portia', 'ppcor', 'grnboost2', 'scenic', 'granie', 'scglue', 'celloracle', 'figr', 'scenicplus'],
 
-  "evaluation_data": "resources/evaluation_datasets/replogle2.h5ad",
-  'consensus': 'resources/prior/replogle2_consensus-num-regulators.json',
-  # "evaluation_data": "resources/evaluation_datasets/frangieh_IFNg_v2.h5ad",
-  # 'consensus': 'resources/prior/frangieh_IFNg_v2_consensus-num-regulators.json',
+  "evaluation_data": "resources/evaluation_datasets/op_perturbation.h5ad",
+  'consensus': 'resources/prior/op_consensus-num-regulators.json',
 
   'layers': ['X'],
 
@@ -29,6 +26,29 @@
   'num_workers': 20,
   'clip_scores': True
 }
+
+if False: # replogle2
+  par = {
+    'reg_type': 'GB',
+    'models_dir': "resources/grn_models/replogle2",
+    'scores_dir': "output/",
+
+    'models': [ 'negative_control', 'positive_control', 'pearson_corr', 'portia', 'grnboost2', 'ppcor', 'scenic'],
+
+
+    "evaluation_data": "resources/evaluation_datasets/replogle2.h5ad",
+    'consensus': 'resources/prior/replogle2_consensus-num-regulators.json',
+
+    'layers': ['X'],
+
+    "tf_all": "resources/prior/tf_all.csv",
+    'skeleton': 'resources/prior/skeleton.csv', 
+    "apply_tf": True,
+    'subsample': -1,
+    'verbose': 4,
+    'num_workers': 20,
+    'clip_scores': True
+  }
 # VIASH END
 
 meta = {
@@ -39,7 +59,6 @@
 sys.path.append(meta["util"])
 
 os.makedirs(par['scores_dir'], exist_ok=True)
-
 # - run consensus 
 from consensus.script import main 
 main(par)