add_dics

.gitignore

@@ -3,6 +3,7 @@
 **/*/tutorial_bulkRNAseq
 **/*/tutorial_spatial
 **/*/tutorial_tldr
+**/*/experiment*
 
 # Byte-compiled / optimized / DLL files
 __pycache__/

dissect/PropsSimulator/simulator.py

@@ -30,7 +30,7 @@ def generate_props(self):
         if not self.config["simulation_params"]["n_samples"]:
             self.config["simulation_params"]["n_samples"] = 1000 * self.n_celltypes
         if not self.config["simulation_params"]["cells_per_sample"]:
-            self.config["simulation_params"]["cells_per_sample"] = 1000
+            self.config["simulation_params"]["cells_per_sample"] = 100
         self.n_sparse = int(self.config["simulation_params"]["n_samples"] * self.config["simulation_params"]["prop_sparse"])
         self.n_complete = self.config["simulation_params"]["n_samples"] - self.n_sparse
 
@@ -111,13 +111,13 @@ def generate_props(self):
         plt.savefig(
             os.path.join(self.simulation_folder, "boxplot_props_complete.pdf")
         )
-        fig = plt.figure()
-        ax = plt.boxplot(self.cells_complete, labels=self.celltypes)
-        plt.ylabel("Count")
-        plt.title("Counts of cell-types in generated samples")
-        plt.savefig(
-            os.path.join(self.simulation_folder, "boxplot_ncells_complete.pdf")
-        )
+        # fig = plt.figure()
+        # ax = plt.boxplot(self.cells_complete, labels=self.celltypes)
+        # plt.ylabel("Count")
+        # plt.title("Counts of cell-types in generated samples")
+        # plt.savefig(
+        #     os.path.join(self.simulation_folder, "boxplot_ncells_complete.pdf")
+        # )
 
         # fig = plt.figure()
         # ax = plt.boxplot(self.props_sparse, labels=self.celltypes)  #
@@ -610,10 +610,10 @@ def simulate(config):
         else:
             sim.simulate(save=True)
     else:
-        print("Number of batches in single-cell data is 1.")
+        print("Number of batches in single-cell data is 1. If this is incorrect, please specify name of the batch column as in the single-cell data object (.obs)")
         sim.simulate(save=True)
     sim.config["simulation_params"]["simulation_folder"] = os.path.join(sim.config["experiment_folder"], "simulation")
     sim.config["simulation_params"]["concentration"] = list(sim.config["simulation_params"]["concentration"])
     sim.config["deconv_params"]["reference"] = os.path.join(sim.config["simulation_params"]["simulation_folder"], "simulated.h5ad")
     save_dict_to_file(sim.config)
-
+

dissect/PropsSimulator/simulator.pyx

@@ -595,7 +595,7 @@ def simulate(config):
         else:
             sim.simulate(save=True)
     else:
-        print("Number of batches in single-cell data is 1.")
+        print("Number of batches in single-cell data is 1. If this is incorrect, please specify name of the batch column as in the single-cell data object (.obs)")
         sim.simulate(save=True)
     sim.config["simulation_params"]["simulation_folder"] = os.path.join(sim.config["experiment_folder"], "simulation")
     sim.config["simulation_params"]["concentration"] = list(sim.config["simulation_params"]["concentration"])

dissect/deconv.py

@@ -4,7 +4,15 @@
 from dissect.dissect_expr import run_dissect_expr
 
 def run_dissect(config):
-        simulate(config)
-        dataset(config)
-        run_dissect_frac(config)
-        run_dissect_expr(config)
+	print("\033[1mStarting simulations\033[0m")
+	simulate(config)
+	print("-----\n")
+
+	print("\033[1mPreprocessing datasets\033[0m")
+	dataset(config)
+	print("-----\n")
+	print("\033[1mEstimating cell type fractions\033[0m")
+	run_dissect_frac(config)
+	print("-----\n")
+	print("\033[1mEstimating cell type-specific gene expression per sample\033[0m")
+	run_dissect_expr(config)

dissect/dissect_expr.py

@@ -33,6 +33,7 @@
 
 
 def run_dissect_expr(config):
+    os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
     test_format = config["deconv_params"]["test_dataset_format"]
     real_path = config["deconv_params"]["test_dataset"]
 
@@ -449,4 +450,4 @@ def call(self, inputs, labels):
 
 if __name__ == "__main__":
     from configs.main_config import config
-    run_dissect_expr(config)
+    run_dissect_expr(config)

dissect/dissect_frac.py

@@ -9,9 +9,11 @@
 import random
 from dissect.utils.utils_fn import normalize_per_batch, reproducibility, ccc_fn
 from sklearn.metrics import mean_squared_error
-
+import logging
 
 def run_dissect_frac(config):
+    os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
+    logging.getLogger('tensorflow').setLevel(logging.FATAL)
     dataset_path = os.path.join(config["experiment_folder"], "datasets")
     if not os.path.exists(dataset_path):
         sys.exit(
@@ -216,7 +218,7 @@ def run_dissect_frac(config):
 
         model_p = tf.keras.models.load_model(model_path)
 
-        print("Running deconvolution")
+        # print("Running deconvolution")
         y_hat = model_p.predict(normalize_per_batch(X_real_test, n_features))
         df_y_hat = pd.DataFrame(y_hat, columns=celltypes)
         df_y_hat.index = sample_names
@@ -234,7 +236,7 @@ def run_dissect_frac(config):
         # df_metrics.to_csv(os.path.join(config["experiment_folder"], "per_step_metrics_{}.txt".format(j)),
         #                sep="\t")
 
-        print("Estimated proportions are saved at {}.".format(results_path))
+        # print("Estimated proportions are saved at {}.".format(results_path))
         if i < 5:
             print("Starting training model {}".format(i))
         j += 1
@@ -243,7 +245,7 @@ def run_dissect_frac(config):
     experiment_path = os.path.join(config["experiment_folder"])
 
     i = 0
-    for i in range(len(config["models"])):
+    for i in range(len(config["deconv_params"]["models"])):
         df_curr = pd.read_table(
             os.path.join(
                 config["experiment_folder"], "dissect_fractions_{}.txt".format(i)
@@ -254,9 +256,9 @@ def run_dissect_frac(config):
             df_ens = df_curr
         else:
             df_ens = df_ens + df_curr
-    df_ens = df_ens / len(config["models"])
+    df_ens = df_ens / len(config["deconv_params"]["models"])
     savepath = os.path.join(config["experiment_folder"], "dissect_fractions.txt")
-    print("Final predictions are saved to {}".format(savepath))
+    print("Predictions are saved to {}".format(savepath))
     df_ens.to_csv(savepath, sep="\t")
 
 if __name__ == "__main__":

docs/index.html

@@ -0,0 +1,19 @@
+<!DOCTYPE html>
+<html lang="en">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>DISSECT v0.1.1</title>
+</head>
+<body>
+    <h1>DISSECT v0.1.1</h1>
+    <nav>
+        <ul>
+            <li><a href="installation.md">Installation</a></li>
+            <!--- <li><a href="expanded_tutorial.html">Step by step tutorial</a></li> -->
+            <li><a href="minimal_tutorial.html">Minimal tutorial</a></li>
+        </ul>
+    </nav>
+</body>
+</html>
+

docs/installation.md

@@ -0,0 +1,37 @@
+# DISSECT
+
+DISSECT is a deep semi-supervised learning framework that performs complete deconvolution of gene expression mixtures such as bulk RNAseq, proteomics and spatial transcriptomics
+
+## Getting Started
+
+These instructions will get you a copy of the project up and running on your local machine for development and testing purposes. Folder tutorials include notebooks to get started.
+
+## Prerequisites
+
+conda >= v22 through [Anaconda](https://docs.anaconda.com/free/anaconda/install/index.html) or [miniconda](https://docs.conda.io/projects/miniconda/en/latest/miniconda-install.html).
+
+
+## Installing
+```
+## Installation
+
+# Create and activate virtual environment
+conda create -y -n dissect python=3.8
+conda activate dissect
+
+# Clone DISSECT
+git clone https://github.com/imsb-uke/DISSECT
+
+# Install dependencies and DISSECT
+
+pip install -r DISSECT/requirements.txt --use-deprecated=legacy-resolver
+pip install DISSECT/.
+
+# Install jupyter lab
+conda install -y jupyter
+
+## Tutorials to get started
+# Go to tutorials directory within DISSECT
+cd DISSECT/tutorials
+
+```