scglue prog

src/control_methods/baseline_corr/script.py

@@ -9,15 +9,15 @@
 
 ## VIASH START
 par = {
-    'multiomics_rna': 'resources/grn-benchmark/multiomics_rna.h5ad',
+    'multiomics_rna': 'resources/grn-benchmark/multiomics_rna_0.h5ad',
     'tf_all': 'resources/prior/tf_all.csv',
-    'causal': False,
+    'causal': True,
     'metacell': False,
     'cell_type_specific': False,
     'impute': False,
     'max_n_links': 50000,
     'corr_method': 'pearson',
-    'prediction': 'resources/grn_models/alldonors_default/pearson.csv',
+    'prediction': 'resources/grn_models/donor_0_default/pearson_causal.csv',
     "seed": 32
 }
 ## VIASH END
@@ -96,9 +96,7 @@ def create_meta_cells(df, n_cells=15):
 tf_all = np.intersect1d(tf_all, gene_names)
 
 print('Noramlize data')
-sc.pp.normalize_total(multiomics_rna)
-sc.pp.log1p(multiomics_rna)
-sc.pp.scale(multiomics_rna)
+multiomics_rna.X = multiomics_rna.layers['lognorm'] 
 
 if par['impute']:
     print("imputing")

src/methods/multi_omics/scglue/main.py

@@ -155,31 +155,23 @@ def run_grn(par):
     np.savetxt(f"{par['temp_dir']}/tfs.txt", tfs, fmt="%s")
 
     #Construct the command 
-    command = ['pyscenic', 'grn', f"{par['temp_dir']}/rna.loom", 
-               f"{par['temp_dir']}/tfs.txt", '-o', f"{par['temp_dir']}/draft_grn.csv", 
-               '--seed', '0', '--num_workers', f"{par['num_workers']}", 
-               '--cell_id_attribute', 'obs_id', '--gene_attribute', 'name']
-    print('Run grn')
-    result = subprocess.run(command,  check=True)
+    # command = ['pyscenic', 'grn', f"{par['temp_dir']}/rna.loom", 
+    #            f"{par['temp_dir']}/tfs.txt", '-o', f"{par['temp_dir']}/draft_grn.csv", 
+    #            '--seed', '0', '--num_workers', f"{par['num_workers']}", 
+    #            '--cell_id_attribute', 'obs_id', '--gene_attribute', 'name']
+    # print('Run grn')
+    # result = subprocess.run(command,  check=True)
 
-    print("Output:")
-    print(result.stdout)
-    print("Error:")
-    print(result.stderr)
+    # print("Output:")
+    # print(result.stdout)
+    # print("Error:")
+    # print(result.stderr)
 
-    if result.returncode == 0:
-        print("Command executed successfully")
-    else:
-        print("Command failed with return code", result.returncode)
-def create_prior(par):
-    atac = ad.read_h5ad(f"{par['temp_dir']}/atac-emb.h5ad")
-    rna = ad.read_h5ad(f"{par['temp_dir']}/rna-emb.h5ad")
-    motif_bed = scglue.genomics.read_bed(par['motif_file'])
-    fs = pd.Index(motif_bed["name"]).intersection(rna.var_names)
+    # if result.returncode == 0:
+    #     print("Command executed successfully")
+    # else:
+    #     print("Command failed with return code", result.returncode)
 
-    atac.var["name"] = atac.var_names
-
-    peaks = atac.var.index
 
     print("Generate TF cis-regulatory ranking bridged by ATAC peaks", flush=True)
     peak_bed = scglue.genomics.Bed(atac.var.loc[peaks])
@@ -311,8 +303,7 @@ def main(par):
     # preprocess(rna, atac, par)
     # print('Train a model', flush=True)
     # training(par)
-    # run_grn(par)
-    create_prior(par)
+    run_grn(par)
     prune_grn(par)
     print('Curate predictions', flush=True)
     pruned_grn = pd.read_csv(

src/metrics/regression_1/script.py

@@ -7,7 +7,7 @@
 par = {
   "perturbation_data": "resources/grn-benchmark/perturbation_data.h5ad",
   "tf_all": "resources/prior/tf_all.csv",
-  "prediction": "resources/grn_models/donor_0_celltype/grnboost2.csv",
+  "prediction": "resources/grn_models/donor_0_default/pearson_causal.csv",
   "method_id": "scenic",
   "min_tf": False,
   "max_n_links": 50000,
@@ -19,9 +19,9 @@
   'max_workers': 4,
 }
 ## VIASH END
-# meta = {
-#   "resources_dir":'src/metrics/regression_1/'
-# }
+meta = {
+  "resources_dir":'src/metrics/regression_1/'
+}
 sys.path.append(meta["resources_dir"])
 from main import main