fixed some naming errors and both model options

app.py

@@ -5,14 +5,15 @@
 
 from composition import composition_server, composition_ui
 from export import export_ui, export_server
-from dgea.dgea_1 import dgea_server, dgea_ui
+from dgea.dgea_scvi_define import dgea_server, dgea_ui
 from tree import tree_server, tree_ui
 
 with open("data/config.json") as f:
     config = json.load(f)
     adata = sc.read_h5ad("data/" + config["adata"])
     tree = pickle.load(open("data/" + config["tree"], "rb")) if "tree" in config else None
     name = config["name"]
+    model_path = config["model_path"]
 
 categorical_columns = adata.obs.select_dtypes(include="category").columns.to_list()
 
@@ -34,6 +35,7 @@ def server(input, output, session):
     _dataframe = reactive.value(adata.obs)
     _adata = reactive.value(adata)
     _tree = reactive.value(tree)
+    _model = reactive.value(model_path)
     composition_server("composition", _dataframe)
     export_server("export")
     dgea_server("dgea", _adata)

data/config.json

@@ -1,5 +1,6 @@
 {
     "adata": "merged.h5ad",
     "tree": "scarches.tree.pkl",
-    "name": "possible_atlas"
+    "name": "possible_atlas",
+    "model_path": "./data"
 }

dgea/dgea_scvi.py

@@ -1,43 +1,30 @@
 import numpy as np
 import pandas as pd
 import scanpy as sc
-import random
 import anndata as ad
-import os
 
-import torch
 from scvi.model import SCANVI, SCVI
 
-def scanvi_dgea(adata:ad.AnnData, groupby:str, reference:str, alternative:str):
-
-    directory_model = "model_scvi_test/"
-    scanvi_subdirectory = os.path.join(directory_model, "scanvi_model")
-
-    try:
-        loaded_model = SCVI.load(directory_model, adata=None)
-    except Exception as e:
-        loaded_model = None
-
-    if isinstance(loaded_model, SCVI):
-        os.makedirs(scanvi_subdirectory, exist_ok=True)
-        scanvi_model = SCANVI.from_scvi_model(loaded_model, unlabeled_category = 'Unknown', labels_key="cell_type")
-        scanvi_model.save(scanvi_subdirectory, overwrite=True)
-        directory_model = scanvi_subdirectory
-        print('is scvi')
+def scanvi_dgea(adata:ad.AnnData, groupby:str, reference:str, alternative:str, directory_model:str):
+
+    if 'cell_type' in adata.obs.columns:
+        model_type = SCANVI
+        print('is scavi')
 
     else:
+        model_type = SCVI
         print('is scanvi')
 
-    SCANVI.prepare_query_anndata(adata = adata, reference_model=directory_model)
+    model_type.prepare_query_anndata(adata = adata, reference_model=directory_model)
 
-    scanvi_model = SCANVI.load_query_data(adata, directory_model)
+    model = model_type.load_query_data(adata, directory_model)
 
     groups = np.array(adata.obs[groupby].unique())
 
     idx1 = adata.obs[groupby] == reference
     idx2 = adata.obs[groupby] == alternative
 
-    dge_change = scanvi_model.differential_expression(adata=adata, groupby=groupby, idx1=idx1, idx2=idx2, mode="change")
+    dge_change = model.differential_expression(adata=adata, groupby=groupby, idx1=idx1, idx2=idx2, mode="change")
 
     epsilon = 1e-10
     dge_change['proba_not_de'] = np.maximum(dge_change["proba_not_de"], epsilon)
@@ -56,6 +43,8 @@ def get_normalized_counts(adata):
     df_counts = pd.DataFrame(dense_matrix, index=adata.obs_names, columns=adata.var_names)
     return df_counts
 
-#adata = sc.read_h5ad('model_scvi_test/adata.h5ad')
-#dge_test = scanvi_dgea(adata, "cell_type", "Endothelial", "Epithelial")
-#print(dge_test.head())
+if __name__ == '__main__':
+    print('Running DGEA test')
+    adata = sc.read_h5ad('/workspaces/SIMBA-Downstream_1/data/atlas.h5ad')
+    dge_test = scanvi_dgea(adata, "cell_type", "Endothelial", "Epithelial", './data')
+    print(dge_test.head())

dgea/run_dgea_scvi.py

@@ -17,6 +17,7 @@ def run_dgea_server(input, output, session,
                     _counts: reactive.Value[pd.DataFrame],
                     _reference: reactive.Value[str],
                     _alternative: reactive.Value[str],
+                    _model: reactive.Value[str],
                     _uniques: reactive.Value[list],
                     _contrast: reactive.Value[str]):
     _category_columns = reactive.value([])
@@ -49,7 +50,8 @@ def handle_run():
         contrast = _contrast.get()
         referece = _reference.get()
         alternative = _alternative.get()
-        run_scanvi(adata, contrast, referece, alternative)
+        model = _model.get()
+        run_scanvi(adata, contrast, referece, alternative, model)
 
     @reactive.effect
     def update_scanvi_result():
@@ -59,8 +61,8 @@ def update_scanvi_result():
 
     @ui.bind_task_button(button_id="run")
     @reactive.extended_task
-    async def run_scanvi(adata, contrast, reference, alternative):
-        dge_change = scanvi_dgea(adata, contrast, reference, alternative)
+    async def run_scanvi(adata, contrast, reference, alternative, model):
+        dge_change = scanvi_dgea(adata, contrast, reference, alternative, model)
 
         counts = get_normalized_counts(adata)
         return dge_change, counts