NID pre onehot

GenNet.py

@@ -58,7 +58,7 @@ def __init__(self):
         self.make_parser_topology(parser_topology)
 
         parser_interpret = subparsers.add_parser("interpret", help="Post-hoc interpretation analysis on the network")
-        self.parser_interpret(parser_interpret)
+        self.make_parser_interpret(parser_interpret)
 
         self.parser = parser
 

GenNet_utils/Interpret.py

@@ -8,9 +8,7 @@
 
 from interpretation.weight_importance import make_importance_values_input
 from interpretation.NID import Get_weight_tsang, GenNet_pairwise_interactions_topn
-from interpretation.RLIPP import calculate_RLIPP
 
-from GenNet_utils.Utility_functions import get_SLURM_id, evaluate_performance
 from GenNet_utils.Train_network import load_trained_network
 
 def interpret(args):
@@ -38,27 +36,35 @@ def get_weight_scores(args):
 
 
 def get_NID_scores(args):
+    print("Interpreting with NID:")
     model, masks = load_trained_network(args)
 
-    if args.layer == "None":
+    print(model.summary())
+
+    if args.layer == None:
         if args.one_hot == 1:
             interp_layer = 3
         else:
             interp_layer = 2
     else:
         interp_layer = args.layer
 
+    print("Interrpeting layer", interp_layer)
     if os.path.exists(args.resultpath + "/NID.csv"):
         print('RLIPP Done')
         interaction_ranking = pd.read_csv(args.resultpath + "/NID.csv")
     else:
+        print("Obtaining the weights")
         w_in, w_out = Get_weight_tsang(model, interp_layer, masks)
+        print("Computing interactions")
+
         interaction_ranking1 = GenNet_pairwise_interactions_topn(w_in[:,1] ,w_out[:,1], masks, n=4)
         interaction_ranking2 = GenNet_pairwise_interactions_topn(w_in[:,0] ,w_out[:,0], masks, n=4)
 
         interaction_ranking = interaction_ranking1.append(interaction_ranking2)
         interaction_ranking = interaction_ranking.sort_values("strength", ascending =False)
         interaction_ranking.to_csv(args.resultpath + "/NID.csv")
+        print("NID results are saved in", args.resultpath + "/NID.csv")
 
     return interaction_ranking
 

GenNet_utils/Train_network.py

@@ -32,6 +32,9 @@ def train_model(args):
     model = None
     masks = None
 
+    args.datapath = args.path
+
+
     if args.genotype_path == "undefined":
         args.genotype_path = args.path
 
@@ -138,8 +141,6 @@ def train_model(args):
     plot_loss_function(args.resultpath)
     model.load_weights(args.resultpath + '/bestweights_job.h5')
     print("Finished")
-
-
     save_train_arguments(args)
 
 
@@ -224,11 +225,10 @@ def train_model(args):
         importance_csv = create_importance_csv(args.datapath, model, masks)
         importance_csv.to_csv(args.resultpath + "connection_weights.csv")
 
-
 def get_network(args):
     """needs the following inputs
         args.inputsize
-        args.l1_value
+        args.L1
         args.L1_act
         args.datapath
         args.genotype_path
@@ -237,40 +237,42 @@ def get_network(args):
     """
     regression = args.regression if hasattr(args, 'regression') else False
 
+    args.L1 = args.regression if hasattr(args, 'regression') else False
+
     if args.network_name == "lasso" and not regression:
         print("lasso network")
-        model, masks = lasso(inputsize=args.inputsize, l1_value=args.l1_value, L1_act=args.L1_act if hasattr(args, 'L1_act') else None)
+        model, masks = lasso(inputsize=args.inputsize, l1_value=args.L1, L1_act=args.L1_act if hasattr(args, 'L1_act') else None)
 
     elif args.network_name == "sparse_directed_gene_l1" and not regression:
         print("sparse_directed_gene_l1 network")
-        model, masks = sparse_directed_gene_l1(inputsize=args.inputsize, l1_value=args.l1_value)
+        model, masks = sparse_directed_gene_l1(inputsize=args.inputsize, l1_value=args.L1)
 
 
     elif args.network_name == "regression_height" and regression:
         print("regression_height network")
-        model, masks = regression_height(inputsize=args.inputsize, l1_value=args.l1_value)
+        model, masks = regression_height(inputsize=args.inputsize, l1_value=args.L1)
 
     elif args.network_name == "gene_network_multiple_filters":
         print("gene_network_multiple_filters network")
         model, masks = gene_network_multiple_filters(datapath=args.datapath, inputsize=args.inputsize, genotype_path=args.genotype_path,
-                                                     l1_value=args.l1_value, L1_act=args.L1_act, 
+                                                     l1_value=args.L1, L1_act=args.L1_act, 
                                                      regression=regression, num_covariates=args.num_covariates,
                                                      filters=args.filters)
 
     elif args.network_name == "gene_network_snp_gene_filters":
         print("gene_network_snp_gene_filters network")
         model, masks = gene_network_snp_gene_filters(datapath=args.datapath, inputsize=args.inputsize, genotype_path=args.genotype_path,
-                                                     l1_value=args.l1_value, L1_act=args.L1_act, 
+                                                     l1_value=args.L1, L1_act=args.L1_act, 
                                                      regression=regression, num_covariates=args.num_covariates,
                                                      filters=args.filters)
     else:
         if os.path.exists(args.datapath + "/topology.csv"):
             model, masks = create_network_from_csv(datapath=args.datapath, inputsize=args.inputsize, genotype_path=args.genotype_path,
-                                                   l1_value=args.l1_value, L1_act=args.L1_act, regression=regression,
+                                                   l1_value=args.L1, L1_act=args.L1_act, regression=regression,
                                                    num_covariates=args.num_covariates)
         elif len(glob.glob(args.datapath + "/*.npz")) > 0:
             model, masks = create_network_from_npz(datapath=args.datapath, inputsize=args.inputsize, genotype_path=args.genotype_path,
-                                                   l1_value=args.l1_value, L1_act=args.L1_act, regression=regression,
+                                                   l1_value=args.L1, L1_act=args.L1_act, regression=regression,
                                                    num_covariates=args.num_covariates, 
                                                    mask_order=args.mask_order if hasattr(args, 'mask_order') else None)
 
@@ -294,7 +296,7 @@ def load_trained_network(args):
 
     get_network needs the following inputs
         args.inputsize
-        args.l1_value
+        args.L1
         args.L1_act
         args.datapath
         args.genotype_path

GenNet_utils/Utility_functions.py

@@ -226,8 +226,8 @@ def save_train_arguments(args, filename="train_args.json"):
 
     # Convert args to a dictionary, taking care of non-serializable types if necessary
     args_dict = vars(args)  # Convert the Namespace to a dictionary
-    with open(filename, 'w') as file:
-        json.dump(args_dict,args.resultpath + file, indent=4)
+    with open(args.resultpath + filename, 'w') as file:
+        json.dump(args_dict, file, cls=NumpyEncoder, indent=4)
 
 
 
@@ -251,3 +251,16 @@ def load_train_arguments(args, filename="train_args.json"):
 
     return args
 
+
+class NumpyEncoder(json.JSONEncoder):
+    def default(self, obj):
+        if isinstance(obj, (np.int_, np.intc, np.intp, np.int8,
+            np.int16, np.int32, np.int64, np.uint8,
+            np.uint16, np.uint32, np.uint64)):
+            return int(obj)
+        elif isinstance(obj, (np.float_, np.float16, np.float32, 
+            np.float64)):
+            return float(obj)
+        elif isinstance(obj, (np.ndarray,)):  # Handle numpy arrays
+            return obj.tolist()
+        return super(NumpyEncoder, self).default(obj)

interpretation/NID.py

@@ -148,7 +148,7 @@ def GenNet_pairwise_interactions_simplified_mask(w_input, w_later, mask):
     interaction_ranking = []
     list_of_combinations = []
     for gene_id in range(mask.shape[1]):
-        neuron_combinations=list(itertools.combinations(mask.row[genemask.col == gene_id], 2 ))
+        neuron_combinations=list(itertools.combinations(mask.row[mask.col == gene_id], 2 ))
 
     for candidate in neuron_combinations:
         strength = (np.minimum(w_input[candidate[0]], w_input[candidate[1]])*w_later[gene_id]).sum()