Merge pull request #10 from Center-for-Health-Data-Science/server_snapshot

Only track whitelisted .yaml configuration files and .py execs are made directly runnable. Fixed DEA.

Author: ValeSora

.gitignore

@@ -48,3 +48,15 @@ venv.bak/
 
 # MacOS
 .DS_Store
+
+# downloaded decoder
+dec.pth
+
+# only keep whitelisted set of config files
+bulkDGD/ioutil/configs/*.yaml
+!model.yaml 
+!pca_scatter.yaml
+!r_values_hist.yaml
+!time_line.yaml
+!one_opt.yaml 
+!two_opt.yaml

bulkDGD/analysis/dea.py

@@ -793,17 +793,15 @@ def get_log2_fold_changes(obs_counts,
     # should return the columns in the order specified by the
     # selection.
     obs_counts = \
-        torch.Tensor(obs_counts.loc[genes_obs].astype("int").values)
-
+        torch.tensor(obs_counts.loc[genes_obs].astype("float").values)
     #-----------------------------------------------------------------#
 
     # Create a tensor containing only those columns containing gene
     # expression data for the predicted mean counts - the 'loc' syntax
     # should return the columns in the order specified by the
     # selection.
     pred_means = \
-        torch.Tensor(pred_means.loc[genes_obs].astype("int").values)
-
+        torch.tensor(pred_means.loc[genes_obs].astype("float").values)
     #-----------------------------------------------------------------#
 
     # Get the mean gene counts for the sample.

bulkDGD/execs/dgd_get_probability_density.py

@@ -369,3 +369,6 @@ def main():
         "were successfully written in " \
         f"'{output_csv_prob_comp_path}'."
     log.info(infostr)
+
+if __name__ == "__main__":
+    main()

bulkDGD/execs/dgd_get_representations.py

@@ -456,3 +456,6 @@ def main():
         "The time data were successfully written in " \
         f"'{output_csv_time_path}'."
     log.info(infostr)
+
+if __name__ == "__main__":
+    main()

bulkDGD/execs/dgd_perform_dea.py

@@ -435,6 +435,10 @@ def main():
         # Get the data frame containing the DEA results for the
         # current sample and the name of the sample.
         df_stats, sample_name = result
+        df_stats["obs_counts"] = obs_counts.loc[sample_name,:]
+        df_stats["dgd_mean"] = pred_means.loc[sample_name,:]
+        df_stats["dgd_r"] = r_values
+        
 
         # Set the path to the output file.
         output_csv_path = \
@@ -465,3 +469,6 @@ def main():
             f"The DEA results for sample '{sample_name}' were " \
             f"successfully written in '{output_csv_path}'."
         log.info(infostr)
+
+if __name__ == "__main__":
+    main()

bulkDGD/execs/dgd_preprocess_samples.py

@@ -337,3 +337,6 @@ def main():
             "was successfully written in " \
             f"'{output_txt_genes_missing_path}'."
         log.info(infostr)
+
+if __name__ == "__main__":
+    main()