Create autonumclusters

BoolODE_gendatasets.py

@@ -0,0 +1,67 @@
+#generating datasets for Boolean models
+
+path_to_boolode = "/home/cbuck016/BoolODE-0.1/"
+
+maxtime = "8"
+numcells = "1000"
+numdatasets = "10"
+
+output_dir = "CuratedData/"
+
+
+# Boolean model 1: mCAD
+model_name = "mCAD"
+output_name = "mCAD"
+list_of_mCAD_sims = ["mCAD-sim-01-ts-800-cells-1000", "mCAD-sim-02-ts-800-cells-1000", "mCAD-sim-03-ts-800-cells-1000", "mCAD-sim-04-ts-800-cells-1000", "mCAD-sim-05-ts-800-cells-1000", "mCAD-sim-06-ts-800-cells-1000", "mCAD-sim-07-ts-800-cells-1000", "mCAD-sim-08-ts-800-cells-1000", "mCAD-sim-09-ts-800-cells-1000", "mCAD-sim-10-ts-800-cells-1000"]
+for mCAD_sim in list_of_mCAD_sims:
+	echo "Simulating "$model_name
+	python3 $path_to_boolode/src/BoolODE.py --path $path_to_boolode/data/$model_name".txt" \
+						--ics $path_to_boolode/data/$model_name"_ics.txt" \
+						--max-time $maxtime --num-cells $numcells \
+						--do-parallel \
+						--outPrefix $output_dir$output_name"/" \
+						--sample-cells 
+
+
+# Boolean model 2: VSC
+model_name = "VSC"
+output_name = "VSC"
+list_of_VSC_sims = ["VSC-sim-01-ts-800-cells-1000", "VSC-sim-02-ts-800-cells-1000", "VSC-sim-03-ts-800-cells-1000", "VSC-sim-04-ts-800-cells-1000", "VSC-sim-05-ts-800-cells-1000", "VSC-sim-06-ts-800-cells-1000", "VSC-sim-07-ts-800-cells-1000", "VSC-sim-08-ts-800-cells-1000", "VSC-sim-09-ts-800-cells-1000", "VSC-sim-10-ts-800-cells-1000"]
+for VSC_sim in list_of_VSC_sims:
+	echo "Simulating "$model_name
+	python3 $path_to_boolode/src/BoolODE.py --path $path_to_boolode/data/$model_name".txt" \
+						--max-time $maxtime --num-cells $numcells \
+						--do-parallel \
+						--outPrefix $output_dir$output_name"/" \
+						--sample-cells 
+
+
+# Boolean model 3: HSC
+model_name = "HSC"
+output_name = "HSC"
+list_of_HSC_sims = ["HSC-sim-01-ts-800-cells-1000", "HSC-sim-02-ts-800-cells-1000", "HSC-sim-03-ts-800-cells-1000", "HSC-sim-04-ts-800-cells-1000", "HSC-sim-05-ts-800-cells-1000", "HSC-sim-06-ts-800-cells-1000", "HSC-sim-07-ts-800-cells-1000", "HSC-sim-08-ts-800-cells-1000", "HSC-sim-09-ts-800-cells-1000", "HSC-sim-10-ts-800-cells-1000"]
+for HSC_sim in list_of_HSC_sims:
+	echo "Simulating "$model_name
+	python3 $path_to_boolode/src/BoolODE.py --path $path_to_boolode/data/$model_name".txt" \
+						--ics $path_to_boolode/data/$model_name"_ics.txt" \
+						--max-time $maxtime --num-cells $numcells \
+						--do-parallel \
+						--outPrefix $output_dir$output_name"/" \
+						--sample-cells 
+
+
+# Boolean model 4: GSD
+model_name = "GSD"
+output_name = "GSD"
+list_of_GSD_sims = ["GSD-sim-01-ts-800-cells-1000", "GSD-sim-02-ts-800-cells-1000", "GSD-sim-03-ts-800-cells-1000", "GSD-sim-04-ts-800-cells-1000", "GSD-sim-05-ts-800-cells-1000", "GSD-sim-06-ts-800-cells-1000", "GSD-sim-07-ts-800-cells-1000", "GSD-sim-08-ts-800-cells-1000", "GSD-sim-09-ts-800-cells-1000", "GSD-sim-10-ts-800-cells-1000"]
+for GSD_sim in list_of_GSD_sims:
+	echo "Simulating "$model_name
+	python3 $path_to_boolode/src/BoolODE.py --path $path_to_boolode/data/$model_name".txt" \
+						--ics $path_to_boolode/data/$model_name"_ics.txt" \
+						--max-time $maxtime --num-cells $numcells \
+						--do-parallel \
+						--outPrefix $output_dir$output_name"/" \
+						--sample-cells 
+
+
+

BoolODE_gendatasets_master.py

@@ -0,0 +1,27 @@
+#BoolODE_gendatasets_master.py
+
+import argparse
+import sys
+import os
+
+pathtoBoolODE = '/home/cbuck016/BoolODE-0.1/'
+outputdir = '/home/cbuck016/BoolODE-0.1/GSD-sims/'
+
+num_simulations = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
+num_cells = [250, 500, 1000, 2000, 5000]
+num_timesteps = [1, 2, 4, 8, 16]
+model_name = 'GSD'
+
+for cells in num_cells:
+    for ts in num_timesteps:
+        for i in num_simulations:
+            experiment_name = model_name + '-ts-' + str(ts) + '00' + '-cells-' + str(cells) + '-sim-' + str(i)
+            input_file_prefix = pathtoBoolODE + 'data/' + model_name 
+            command = 'python3 ' + pathtoBoolODE + 'src/BoolODE.py --path ' + input_file_prefix + '.txt ' \
+            + ' --ics ' + input_file_prefix + '_ics.txt ' \
+            + ' --max-time ' + str(ts) + ' --num-cells ' + str(cells) \
+            + ' --do-parallel ' \
+            + ' --outPrefix ' + outputdir + experiment_name + "/sim-" + str(i) + "-" \
+            + ' --sample-cells'
+            print(command)
+            os.system(command)

autonumclusters.py

@@ -0,0 +1,89 @@
+#Silhouette Model example code below retrieved from scitkit-learn, currently editing to fit/run on BoolODE expression data
+
+from sklearn.cluster import KMeans
+from sklearn.metrics import silhouette_samples, silhouette_score
+
+import os
+import sys
+import pandas as pd
+import matplotlib.pyplot as plt
+import matplotlib.cm as cm
+import numpy as np
+import matplotlib.style as style
+from optparse import OptionParser
+from pandas import DataFrame
+
+def parseArgs(args):
+    parser = OptionParser()
+    parser.add_option('', '--expressionfile', type='str', help='Path to ExpressionData.csv file')
+    parser.add_option('', '--outPrefix', type='str', default='', help='Prefix for output files')
+
+    (opts, args) = parser.parse_args(args)
+    return opts, args
+
+
+def main(args):
+    opts, args = parseArgs(args)
+    expressionfile = opts.expressionfile
+    outPrefix = opts.outPrefix
+    if expressionfile is None or len(expressionfile) == 0:
+        print("Please specify path to ExpressionData.csv file")
+        sys.exit
+    if len(expressionfile) > 0:
+        expfileDF = pd.read_csv(expressionfile, index_col=0)
+        data = expfileDF.transpose()
+        print(expfileDF.shape)
+        print(data.shape)
+    if len(outPrefix) > 0:
+        if '/' in outPrefix:
+            outDir = '/'.join(outPrefix.split('/')[:-1])
+            if not os.path.exists(outDir):
+                print(outDir, "does not exist, creating it...")
+                os.makedirs(outDir)
+
+        range_n_clusters = [2, 3, 4, 5, 6, 7, 8, 9, 10]
+        silhouette_avg_n_clusters = []
+
+        for n_clusters in range_n_clusters:
+            clusterer = KMeans(n_clusters=n_clusters, random_state=42)
+            cluster_labels = clusterer.fit_predict(data)
+
+            silhouette_avg = silhouette_score(data, cluster_labels)
+            print("For n_clusters =", n_clusters, "The average silhouette_score is :", silhouette_avg)
+
+            silhouette_avg_n_clusters.append(silhouette_avg)
+
+        best_avg_silhouette_value = max(silhouette_avg_n_clusters)
+        best_num_cluster = silhouette_avg_n_clusters.index(best_avg_silhouette_value)
+
+        if index_best_num_cluster == 0:
+           best_num_cluster = 2
+        elif index_best_num_cluster == 1:
+           best_num_cluster = 3
+        elif index_best_num_cluster == 2:
+           best_num_cluster = 4
+        elif index_best_num_cluster == 3:
+           best_num_cluster = 5
+        elif index_best_num_cluster == 4:
+           best_num_cluster = 6
+        elif index_best_num_cluster == 5:
+           best_num_cluster = 7
+        elif index_best_num_cluster == 6:
+           best_num_cluster = 8
+        elif index_best_num_cluster == 7:
+           best_num_cluster = 9
+        elif index_best_num_cluster == 8:
+           best_num_cluster = 10
+        else:
+           print("No best cluster number found")
+
+        print("The best average silhouette score is: ", best_avg_silhouette_value)
+
+        df = pd.DataFrame({'Average Silhouette Method':[expressionfile, best_num_cluster, best_avg_silhouette_value]}, index=['file name', 'predicted number of clusters', 'average silhouette score'])
+        print(df)
+        df.to_csv(outPrefix + 'silhouettescores.csv')
+
+
+if __name__ == "__main__":
+    main(sys.argv)
+

autonumclusters_master.py

@@ -0,0 +1,19 @@
+#Silhouette Method for use with output results from BoolODE datasets
+
+import os
+
+pathtoBoolOutFile = '/home/cbuck016/BoolODE-0.1/mCAD-sims/'
+
+num_simulations = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
+num_cells = [250, 500, 1000, 2000, 5000]
+num_timesteps = [1, 2, 4, 8, 16]
+model_name = 'mCAD'
+
+for cells in num_cells:
+  for ts in num_timesteps:
+    for i in num_simulations:
+        experiment_name = model_name + '-ts-' + str(ts) + '00' + '-cells-' + str(cells) + '-sim-' + str(i)
+        command = 'python3 ' + 'autonumclusters.py --expressionfile ' + pathtoBoolOutFile + experiment_name \
+        + '/' + 'sim-' + str(i) + '-ExpressionData.csv ' + '--outPrefix ' + pathtoBoolOutFile + experiment_name + "/" 
+        print(command)
+        os.system(command)