scripts to generate jitter plot of simulation scores

This commit includes a python script and an update to the tally job for generating jitter/strip plots of the simulation StrainID scores.

scripts/make_jitter.py
-seaborn library-based plot for showing spread of scores assigned to each strain for each simulation "experiment" (synthetic_strain x depth)
job/tally_results.sh
-PBS script updated with py script calls to generate figs

Author: owlang

paper/SyntheticStrain/job/tally_results.sh

@@ -36,3 +36,8 @@ do
 		python $RUNTIME -i <(grep 'real' $DIR/ID/*.time) > $DIR\_runtimes.txt
 	done
 done
+
+# Make figs
+JITTER=scripts/make_jitter.py
+python $JITTER -i <(awk '{print "results/"$1"_"$2"_scores.txt"}' depth_simulations.txt |grep 'sacCer3') -t Fig4B
+python $JITTER -i <(awk '{print "results/"$1"_"$2"_scores.txt"}' depth_simulations.txt |grep 'hg19') -t Fig4C

paper/SyntheticStrain/scripts/make_jitter.py

@@ -0,0 +1,81 @@
+from os import listdir
+from os.path import isfile, join
+import sys
+import re
+import random
+import argparse
+import matplotlib.pyplot as plt
+import numpy as np
+import seaborn as sns
+
+# Python 3.6+
+# relies on dict insertion order
+
+# Check Seaborn documentation: https://seaborn.pydata.org/generated/seaborn.swarmplot.html
+
+def getParams():
+	'''Parse parameters from the command line'''
+	parser = argparse.ArgumentParser(description='')
+
+	parser.add_argument('-i','--input-files', metavar='file_list', dest='file_list', required=True, help='Script takes text file with a list of file paths to the parsed results of each simulation experiment "<genome>_<strain>\t<depth>\t..." (i.e. "depth_simulations.txt")')
+	parser.add_argument('-t','--title', metavar='figure_title', dest='title', required=True, help='Title to add to figure which is also used to name output image file')
+
+	args = parser.parse_args()
+	return(args)
+
+def parse_data(data_file):
+	'''Parse the data file with the simulation results (<genome>_<strain>_<depth>_scores.txt)'''
+	data = {"value":[],"strain":[]}
+	index_keys = []
+	reader = open(data_file,'r')
+	for line in reader:
+		tokens = line.strip().split("\t")
+		if(tokens[0].find("#")==0):
+			# Initialize strain keys
+			index_keys = tokens[1:]
+			for strain in index_keys: data.update({strain:[]})
+			continue
+		for i in range(len(index_keys)):
+			data["strain"].append(index_keys[i])
+			# Set visualize-able values for "Inf" and "NaN"
+			if(tokens[i+1]=="Inf"):
+				data["value"].append(20.0)
+			elif(tokens[i+1]=="NaN"):
+				data["value"].append(-10.0)
+			else:
+				# Parse as float
+				data["value"].append(float(tokens[i+1]))
+	reader.close()
+
+	return(data)
+
+if __name__ == "__main__":
+	'''Plot jitter/stripplot in R fashion using seaborn library'''
+	args = getParams()
+	# Initialize variables
+	SIZE = 2
+	JITTER = 1
+	all_data = {"experiment":[],"value":[],"strain":[]}
+	# Parse list of file paths
+	i_reader = open(args.file_list,'r')
+	for line in i_reader:
+		data_file = line.strip().split("\t")[0]
+		# Parse datafile for values
+		data = parse_data(data_file)
+		# Merge values into master dataframe
+		all_data["experiment"].extend([data_file]*len(data["value"]))
+		all_data["value"].extend(data["value"])
+		all_data["strain"].extend(data["strain"])
+	i_reader.close()
+
+	# 1000 x 2 x 6 = 12000 points to plot
+	# Plot data points using seaborn and label axes
+	ax = sns.stripplot(x="experiment", y="value", hue="strain", data=all_data, size=SIZE, jitter=JITTER)
+	ax.set_ylabel("log2 score")
+	ax.set_xlabel("Simulation Experiment")
+	ax.set_xticklabels(ax.get_xticklabels(),rotation = 90)
+	# View/save plot
+	#plt.show()
+	out_png_fn = args.title.replace(" ","_")+".svg"
+	plt.savefig(out_png_fn)
+	#print(out_png_fn)