cpgtools 2.0.5__py3-none-any.whl

cpgtools-2.0.5.data/scripts/beta_profile_region.py

@@ -0,0 +1,152 @@
+#!python
+
+"""
+Description
+-----------
+This program calculates methylation profile (i.e., average
+beta value) around user specified genomic regions.
+
+Example of input BED6+ file
+---------------------------
+chr22   44021512        44021513        cg24055475      0.9231  -
+chr13   111568382       111568383       cg06540715      0.1071  +
+chr20   44033594        44033595        cg21482942      0.6122  -
+
+Example of input BED3+ file
+---------------------------
+chr1    15864   15865
+chr1    18826   18827
+chr1    29406   29407
+"""
+
+
+import sys,os
+import collections
+import subprocess
+import numpy as np
+from optparse import OptionParser
+from cpgmodule._version import __version__
+from cpgmodule import ireader
+from cpgmodule.utils import *
+from cpgmodule import BED
+
+__author__ = "Liguo Wang"
+__copyright__ = "Copyleft"
+__credits__ = []
+__license__ = "GPL"
+__maintainer__ = "Liguo Wang"
+__email__ = "wang.liguo@mayo.edu"
+__status__ = "Development"
+
+def main():
+	
+	usage="%prog [options]" + "\n"
+	parser = OptionParser(usage,version="%prog " + __version__)
+	parser.add_option("-i","--input_file",action="store",type="string",dest="input_file",help="BED6+ file specifying the C position. This BED file should have at least six columns (Chrom, ChromStart, ChromeEnd, Name, Beta_value, Strand). BED6+ file can be a regular text file or compressed file (.gz, .bz2).")
+	parser.add_option("-r","--region",action="store",type="string",dest="region_file",help="BED3+ file of genomic regions. This BED file should have at least three columns (Chrom, ChromStart, ChromeEnd). If the 6-th column does not exist, all regions will be considered as on \"+\" strand. ")
+	parser.add_option("-d","--downstream",action="store",type="int",dest="downstream_size",default=2000,help="Size of extension to downstream. default=%default (bp)")
+	parser.add_option("-u","--upstream",action="store",type="int",dest="upstream_size",default=2000,help="Size of extension to upstream. default=%default (bp)")
+	parser.add_option("-o","--output",action="store",type='string', dest="out_file",help="The prefix of the output file.")
+	(options,args)=parser.parse_args()
+	
+	print ()
+
+	if not (options.input_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(101)
+
+	if not (options.region_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(102)
+				
+	if not (options.out_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(103)	
+	
+	FOUT = open(options.out_file + '.txt','w')
+	ROUT = open(options.out_file + '.r','w')
+	print ("\t".join(["Group","Relative_position(5'->3')", "Average_beta"]), file=FOUT)
+	
+	#step1: read CpG file
+	printlog("Reading CpG file: \"%s\"" % (options.input_file))
+	cpg_ranges = read_CpG_bed(options.input_file)
+		
+	#step2: read region file
+	printlog("Reading BED file: \"%s\"" % (options.region_file))
+	
+	region_list = []
+	for chrom, st, end, strand in read_region_bed(options.region_file):
+		region_list.append((chrom, st, end, strand))
+	region_list = list(set(region_list))
+
+	printlog("Calculate average beta ...")
+	s = coverage_over_range(region_list,cpg_ranges)
+	for i in sorted(s):
+		print ('\t'.join(["User_region", str(i), str(s[i])]), file=FOUT)	
+	print ('User_region <- c(%s)' % ','.join([str(s[i]) for i in sorted(s)]), file=ROUT)
+	user_region_datapoints = len(s)
+	
+	if options.upstream_size > 0:
+		printlog("Get upstream regions of  \"%s\"" % (options.region_file))
+		upstream_region = []
+		for (chrom, st, end, strand) in region_list:
+			if strand == '+':
+				upstream_st = max(st - options.upstream_size, 0)
+				upstream_end = st
+				upstream_region.append((chrom, upstream_st, upstream_end, strand))
+			elif strand == '-':
+				upstream_st = end
+				upstream_end = end + options.upstream_size
+				upstream_region.append((chrom, upstream_st, upstream_end, strand))
+		upstream_region = list(set(upstream_region))
+
+	s = coverage_over_range(upstream_region,cpg_ranges)
+	for i in sorted(s):
+		print ('\t'.join(["Upstream_region", str(i), str(s[i])]), file=FOUT)	
+	print ('Upstream_region <- c(%s)' % ','.join([str(s[i]) for i in sorted(s)]), file=ROUT)
+	upstream_datapoints = len(s)
+
+
+	if options.downstream_size > 0:
+		printlog("Get downstream regions of  \"%s\"" % (options.region_file))
+		downstream_region = []
+		for (chrom, st, end, strand) in region_list:
+			if strand == '+':
+				downstream_st = end
+				downstream_end = end + options.downstream_size
+				downstream_region.append((chrom, downstream_st, downstream_end, strand))
+			elif strand == '-':
+				downstream_st = st
+				downstream_end = max(st - options.downstream_size, 0)
+				downstream_region.append((chrom, downstream_st, downstream_end, strand))
+		downstream_region = list(set(downstream_region))
+	s = coverage_over_range(downstream_region,cpg_ranges)
+	for i in sorted(s):
+		print ('\t'.join(["Downstream_region", str(i), str(s[i])]), file=FOUT)	
+	print ('Downstream_region <- c(%s)' % ','.join([str(s[i]) for i in sorted(s)]), file=ROUT)
+	downstream_datapoints = len(s)
+	
+	total_datapoints = upstream_datapoints + downstream_datapoints + user_region_datapoints
+	print('\n')
+	print ('pdf(file=\"%s\", width=6, height=6)' % (options.out_file + '.pdf'),file=ROUT)
+	print ('plot(0:%d, c(Upstream_region, User_region, Downstream_region),ylim=c(0,1), xaxt="n",xlab="", ylab="Average methylation", type="l", col="red")' % (total_datapoints -1), file=ROUT)
+	print ('abline(v = c(%d,%d),col="blue", lty="dashed")' % (upstream_datapoints-1, upstream_datapoints + user_region_datapoints - 1), file=ROUT)
+	print ('abline(h = 0.5,col="grey", lty="dashed")', file=ROUT)
+	print ('text(x=c(%d, %d), y=0.9, cex=0.7, labels=c("Upstream\\n(5\'->3\')","Downstream\\n(5\'->3\')"))' % (50, total_datapoints - 50), file=ROUT)
+	print ('dev.off()',file=ROUT)	
+	
+	FOUT.close()
+	ROUT.close()
+	try:
+		subprocess.call("Rscript " + options.out_file + '.r', shell=True)
+	except:
+		print ("Cannot generate pdf file from " + options.out_file + '.r', file=sys.stderr)
+		pass        
+		
+		
+
+if __name__=='__main__':
+	main()	

cpgtools-2.0.5.data/scripts/beta_selectNBest.py

@@ -0,0 +1,116 @@
+#!python
+
+"""
+#=========================================================================================
+Select the K best features according to the K highest scores. Scores can be measured by:
+
+* ANOVA F-value between label/feature for classification tasks.
+* Mutual information for a discrete target.	
+* Chi-squared stats of non-negative features for classification tasks.
+
+Example of input data file
+---------------------------
+CpG_ID	Sample_01	Sample_02	Sample_03	Sample_04
+cg_001	0.831035	0.878022	0.794427	0.880911
+cg_002	0.249544	0.209949	0.234294	0.236680
+cg_003	0.845065	0.843957	0.840184	0.824286
+"""
+import sys
+import numpy as np
+from optparse import OptionParser
+from cpgmodule._version import __version__
+from cpgmodule.utils import *
+import pandas as pd
+
+from sklearn.feature_selection import SelectKBest
+from sklearn.feature_selection import chi2,f_classif,mutual_info_classif
+
+__author__ = "Liguo Wang"
+__copyright__ = "Copyleft"
+__credits__ = []
+__license__ = "GPL"
+__maintainer__ = "Liguo Wang"
+__email__ = "wang.liguo@mayo.edu"
+__status__ = "Development"
+
+	
+def main():
+	
+	usage="%prog [options]" + "\n"
+	parser = OptionParser(usage,version="%prog " + __version__)
+	parser.add_option("-i","--input_file",action="store",type="string",dest="input_file",help="Tab-separated data frame file containing beta values with the 1st row containing sample IDs and the 1st column containing CpG IDs.")
+	parser.add_option("-g","--group",action="store",type="string",dest="group_file",help="Comma-separated group file defining the biological groups of each sample.")
+	parser.add_option("-c","--topK",action="store",type='int', dest="cpg_count", default=100, help="Number of top features to select. default=%default" )
+	parser.add_option("-s","--score-function",action="store",type='string', dest="score_function", default='chisq', help="Scoring function used to measure the dependency between features scores and labels. Must be \"chisq\" (chi-squared statistic), \"anova\" (ANOVA F-value), or \"mi\" (mutual information). default=%default" )
+	parser.add_option("-o","--output",action="store",type='string', dest="out_file",help="The prefix of the output file.")
+	(options,args)=parser.parse_args()
+	
+	print ()
+	if not (options.input_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(101)
+	if not (options.group_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(102)	
+	if not (options.out_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(103)	
+	
+	printlog("Reading input file: \"%s\"" % (options.input_file))
+	df1 = pd.read_csv(options.input_file, index_col = 0, sep="\t")
+	#print (df1)
+	
+	#remove any rows with NAs
+	df2 = df1.dropna(axis=0, how='any')
+	printlog("%d rows with missing values were removed." % (len(df1) - len(df2)))
+	#print (df2)
+	
+	printlog("Transposing data matrix ... ")
+	df2 = df2.T
+	total_feature = len(df2.columns)
+	printlog("Total number of features: %d " % (total_feature))
+	#print (df2)
+	
+	
+	printlog("Reading group file: \"%s\"" % (options.group_file))
+	group = pd.read_csv(options.group_file, index_col=0, header=0,names=['Sample_ID', 'Group_ID'])
+	a = pd.Series(list(group['Group_ID']))	#a is *string labels* for groups: ['Normal', 'Normal', 'Normal', 'Normal', 'Normal', 'Cancer', 'Cancer', 'Cancer', 'Cancer']
+	#print (a)
+	y, tmp = pd.factorize(a)	#y is *numeric labels* for groups: [0 0 0 0 0 1 1 1 1]
+	#print (np.array(y))
+	
+	if options.cpg_count < total_feature:
+		
+		if options.score_function == 'anova':
+			printlog ("Using ANOVA F value to select features ...")
+			selector = SelectKBest(f_classif, k = options.cpg_count)
+		elif options.score_function == 'mi':
+			printlog ("Using Mutual Information to select features ...")
+			selector = SelectKBest(mutual_info_classif, k = options.cpg_count)
+		elif options.score_function == 'chisq':
+			printlog ("Using Chi Square statistic to select features ...")
+			selector = SelectKBest(chi2, k = options.cpg_count)
+		else:
+			printlog("Unknown function: %s" % options.score_function)
+			sys.exit(0)
+	else:
+		printlog("Doing nothing! '-k' >= the total number of features in \"%s\"" % (options.input_file))
+		sys.exit(0)
+	
+	
+	selector.fit_transform(df2, np.array(y))
+	cols = selector.get_support(indices=False)
+	selected_data = df2.loc[:,cols]
+	selected_featureNum = len(selected_data.columns)
+	printlog("Total number of selected features : %d " % (selected_featureNum))
+	#print (selected_data)
+	
+	printlog("Writing to file: \"%s\"" % (options.out_file + '.selectedFeatures.tsv'))
+	pd.DataFrame.to_csv(selected_data.T, options.out_file + '.selectedFeatures.tsv', sep="\t", index_label="sample")
+	
+
+if __name__=='__main__':
+	main()	

cpgtools-2.0.5.data/scripts/beta_stacked_barplot.py

@@ -0,0 +1,119 @@
+#!python
+
+"""
+Description
+-----------
+This program creates stacked barplot for each sample. The stacked barplot showing
+the proportions of CpGs whose beta values are falling into these 4 ranges:
+ * [0.00,  0.25] 	#first quantile
+ * [0.25,  0.50]	#second quantile
+ * [0.50,  0.75]	#third quantile
+ * [0.75,  1.00]	#forth quantile
+
+Example of input data file
+---------------------------
+CpG_ID	Sample_01	Sample_02	Sample_03	Sample_04
+cg_001	0.831035	0.878022	0.794427	0.880911
+cg_002	0.249544	0.209949	0.234294	0.236680
+#=========================================================================================
+
+Note: Please name your sample IDs using only "letters" [a-z, A-Z], "numbers" [0-9], and "_"; and
+your sample ID should start with a letter. 
+
+"""
+
+
+import sys,os
+import collections
+import subprocess
+import numpy as np
+from optparse import OptionParser
+from cpgmodule._version import __version__
+from cpgmodule import ireader
+from cpgmodule.utils import *
+from cpgmodule import BED
+import pandas as pd
+
+__author__ = "Liguo Wang"
+__copyright__ = "Copyleft"
+__credits__ = []
+__license__ = "GPL"
+__maintainer__ = "Liguo Wang"
+__email__ = "wang.liguo@mayo.edu"
+__status__ = "Development"
+
+
+def quarter_count(lst):
+	"""
+	count number of beta values falling into each quarter
+	Note: beta >= 0 and beta <=1
+	"""
+	q1 = 0
+	q2 = 0
+	q3 = 0
+	q4 = 0
+	for i in lst:
+		try:
+			j = float(i)
+		except:
+			continue
+		if not isinstance(j, float):
+			continue
+		if j < 0:
+			continue
+		elif j <= 0.25:
+			q1 += 1
+		elif j <= 0.50:
+			q2 += 1
+		elif j <= 0.75:
+			q3 += 1
+		elif j <= 1:
+			q4 += 1
+		else:
+			continue
+	return [q1, q2, q3, q4]
+	
+def main():
+	
+	usage="%prog [options]" + "\n"
+	parser = OptionParser(usage,version="%prog " + __version__)
+	parser.add_option("-i","--input_file",action="store",type="string",dest="input_file",help="Data frame file containing beta values with the 1st row containing sample IDs and the 1st column containing CpG IDs.")
+	parser.add_option("-o","--output",action="store",type='string', dest="out_file",help="The prefix of the output file.")
+	(options,args)=parser.parse_args()
+	
+	print ()
+	if not (options.input_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(101)
+	
+	if not (options.out_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(103)	
+	
+	printlog("Reading beta file: \"%s\"" % (options.input_file))
+	data = pd.read_csv(options.input_file,sep='\t')
+	samples = data.columns[1:]
+
+	ROUT = open(options.out_file + '.r','w')
+	print ('pdf(file=\"%s\", width=10, height=10)' % (options.out_file + '.pdf'),file=ROUT)
+
+	for s in samples:
+		tmp = quarter_count(data[s])
+		print ('%s <- c(%s)' % (s, ','.join([str(i) for i in tmp])), file=ROUT)
+	print ("cc = rev(c('#d7191c', '#fdae61', '#a6d96a', '#1a9641'))", file=ROUT)
+	print ('legend = c("beta [0.00 - 0.25]", "beta [0.25 - 0.50]", "beta [0.50 - 0.75]", "beta [0.75 - 1.00]")', file=ROUT)
+	print ('nm = c(%s)' % ','.join(['"' + s + '"' for s in samples]), file=ROUT)
+	print ('barplot(cbind(%s), col = cc, names.arg = nm, cex.names = 0.8, ylab = "Percentage", ylim=c(0,119), las=2, legend.text = legend)' % (','.join([s + ' * 100/sum(' + s + ')' for s in samples])), file=ROUT)
+	ROUT.close()
+	
+	try:
+		subprocess.call("Rscript " + options.out_file + '.r', shell=True)
+	except:
+		print ("Cannot generate pdf file from " + options.out_file + '.r', file=sys.stderr)
+		pass        	
+
+if __name__=='__main__':
+	main()	
+	

cpgtools-2.0.5.data/scripts/beta_stats.py

@@ -0,0 +1,101 @@
+#!python
+
+"""
+Description
+-----------
+This program gives basic information on CpGs located in each genomic region. 
+It adds six columns to the input BED file:
+ 1. Number of CpGs detected in the genomic region
+ 2. Min methylation level
+ 3. Max methylation level
+ 4. Average methylation level across all CpGs
+ 5. Median methylation level across all CpGs
+ 6. Standard deviation
+ 
+Example of input BED6+ file
+---------------------------
+chr22   44021512        44021513        cg24055475      0.9231  -
+chr13   111568382       111568383       cg06540715      0.1071  +
+chr20   44033594        44033595        cg21482942      0.6122  -
+"""
+
+
+import sys,os
+import collections
+import subprocess
+import numpy as np
+from optparse import OptionParser
+from cpgmodule._version import __version__
+from cpgmodule import ireader
+from cpgmodule.utils import *
+from cpgmodule import BED
+
+__author__ = "Liguo Wang"
+__copyright__ = "Copyleft"
+__credits__ = []
+__license__ = "GPL"
+__maintainer__ = "Liguo Wang"
+__email__ = "wang.liguo@mayo.edu"
+__status__ = "Development"
+
+def main():
+	usage="%prog [options]" + "\n"
+	parser = OptionParser(usage,version="%prog " + __version__)
+	parser.add_option("-i","--input_file",action="store",type="string",dest="input_file",help="BED6+ file specifying the C position. This BED file should have at least six columns (Chrom, ChromStart, ChromeEnd, Name, Beta_value, Strand).  Note: the first base in a chromosome is numbered 0. This file can be a regular text file or compressed file (.gz, .bz2)")
+	parser.add_option("-r","--region",action="store",type="string",dest="region_file",help="BED3+ file of genomic regions. This BED file should have at least 3 columns (Chrom, ChromStart, ChromeEnd).")
+	parser.add_option("-o","--output",action="store",type='string', dest="out_file",help="The prefix of the output file.")
+	(options,args)=parser.parse_args()
+	
+	print ()
+
+	if not (options.input_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(101)
+
+	if not (options.region_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(102)
+				
+	if not (options.out_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(103)	
+	
+	FOUT = open(options.out_file + '.txt','w')
+	
+	#step1: read CpG file
+	printlog("Reading CpG file: \"%s\"" % (options.input_file))
+	cpg_ranges = read_CpG_bed(options.input_file)
+		
+	#step2: read region file
+	printlog("Reading BED file: \"%s\"" % (options.region_file))
+	
+	printlog("Writing to: \"%s\"" % (options.out_file + '.txt'))
+	region_list = []
+	for l in ireader.reader(options.region_file):
+		if l.startswith('#'):
+			continue
+		if l.startswith('track'):
+			continue
+		if l.startswith('browser'):
+			continue
+		f = l.split()
+		if len(f) < 3:
+			continue
+		try:
+			chrom = f[0]
+			st = int(f[1])
+			end = int(f[2])
+		except:
+			print (l + '\t' + '\t'.join(['NA']*6, file=FOUT))
+			continue
+		tmp = stats_over_range(cpg_ranges, chrom, st, end)
+		print (l + '\t' + '\t'.join([str(i) for i in tmp]), file=FOUT)		
+	
+	FOUT.close()
+
+if __name__=='__main__':
+	main()	
+				

cpgtools-2.0.5.data/scripts/beta_tSNE.py

@@ -0,0 +1,179 @@
+#!python
+
+"""
+Description
+-----------
+This program performs t-SNE (t-Distributed Stochastic Neighbor Embedding) analysis for samples.
+
+Example of input data file
+---------------------------
+CpG_ID	Sample_01	Sample_02	Sample_03	Sample_04
+cg_001	0.831035	0.878022	0.794427	0.880911
+cg_002	0.249544	0.209949	0.234294	0.236680
+cg_003	0.845065	0.843957	0.840184	0.824286
+
+Example of the input group file
+---------------------------
+Sample,Group
+Sample_01,normal
+Sample_02,normal
+Sample_03,tumor
+Sample_04,tumor
+
+Notes
+-----
+* Rows with missing values will be removed
+* Beta values will be standardized into z scores
+* Only the first two components will be visualized
+* Different perplexity values can result in significantly different results
+"""
+
+
+import sys
+import subprocess
+from optparse import OptionParser
+from cpgmodule.utils import *
+from cpgmodule._version import __version__
+import pandas as pd
+from sklearn.preprocessing import StandardScaler
+from sklearn.manifold import TSNE
+
+__author__ = "Liguo Wang"
+__copyright__ = "Copyleft"
+__credits__ = []
+__license__ = "GPL"
+__maintainer__ = "Liguo Wang"
+__email__ = "wang.liguo@mayo.edu"
+__status__ = "Development"
+
+def pick_colors(n):
+	my_colors = [
+	"#F0A3FF", "#0075DC", "#993F00", "#4C005C", "#191919", "#005C31", "#2BCE48", "#FFCC99", "#808080", "#94FFB5", "#8F7C00", "#9DCC00", "#C20088", "#003380", "#FFA405", "#FFA8BB", "#426600", "#FF0010", "#5EF1F2", "#00998F", "#E0FF66", "#740AFF", "#990000", "#FFFF80", "#FFE100", "#FF5005"]
+	if n > len(my_colors):
+		print ("Only support 26 different colors", file = sys.stderr)
+		sys.exit()
+	return my_colors[0:n]
+
+def main():
+	
+	usage="%prog [options]" + "\n"
+	parser = OptionParser(usage,version="%prog " + __version__)
+	parser.add_option("-i","--input_file",action="store",type="string",dest="input_file",help="Tab-separated data frame file containing beta values with the 1st row containing sample IDs and the 1st column containing CpG IDs.")
+	parser.add_option("-g","--group",action="store",type="string",dest="group_file",help="Comma-separated group file defining the biological groups of each sample. Different groups will be colored differently in the t-SNE plot. Supports a maximum of 20 groups.")
+	parser.add_option("-p","--perplexity",action="store",type='int', dest="perplexity_value", default=5, help="This is a tunable parameter of t-SNE, and has a profound effect on the resulting 2D map. Consider selecting a value between 5 and 50, and the selected value should be smaller than the number of samples (i.e., number of points on the t-SNE 2D map). Default = %default" )
+	parser.add_option("-n","--ncomponent",action="store",type='int', dest="n_components", default=2, help="Number of components. default=%default" )
+	parser.add_option("--n_iter",action="store",type='int', dest="n_iterations", default=5000, help="The maximum number of iterations for the optimization. Should be at least 250. default=%default" )
+	parser.add_option("--learning_rate",action="store",type='float', dest="learning_rate", default=200.0, help="The learning rate for t-SNE is usually in the range [10.0, 1000.0]. If the learning rate is too high, the data may look like a ‘ball’ with any point approximately equidistant from its nearest neighbors. If the learning rate is too low, most points may look compressed in a dense cloud with few outliers. If the cost function gets stuck in a bad local minimum increasing the learning rate may help. default=%default" )
+	parser.add_option("-l","--label",action="store_true",default=False,dest="text_label",help="If True, sample ids will be added underneath the data point. default=%default")
+	parser.add_option("-c","--char",action="store",type='int', default=1, dest="plot_char",help="Ploting character: 1 = 'dot', 2 = 'circle'. default=%default")
+	parser.add_option("-a","--alpha",action="store",type='float', default=0.5, dest="plot_alpha",help="Opacity of dots. default=%default")
+	parser.add_option("-x","--loc",action="store",type='int', default=1, dest="legend_location",help="Location of legend panel: 1 = 'topright', 2 = 'bottomright', 3 = 'bottomleft', 4 = 'topleft'. default=%default")
+	parser.add_option("-o","--output",action="store",type='string', dest="out_file",help="The prefix of the output file.")
+	
+	(options,args)=parser.parse_args()
+	
+	#print (options.text_label)
+	#sys.exit(0)
+	print ()
+	if not (options.input_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(101)
+	
+	if not (options.out_file):
+		print (__doc__)
+		parser.print_help()
+		sys.exit(103)	
+	if options.n_components < 2:
+		options.n_components = 2
+		
+	pch = {1:20, 2:1}
+	legend_pos = {1:'topright', 2: 'bottomright', 3:'bottomleft', 4:'topleft'}
+
+	printlog("Reading input file: \"%s\" ..." % (options.input_file))
+	df1 = pd.read_csv(options.input_file, index_col = 0, sep="\t")
+	
+	n_samples = df1.shape[1]
+	#print (n_samples)
+	if (options.perplexity_value > n_samples):
+		options.perplexity_value = int(n_samples/2)
+		printlog("Perplexigty value is set to %d" % options.perplexity_value)
+	
+	#remove NA and transpose
+	df2 = df1.dropna(axis=0, how='any').T
+	printlog("%d rows with missing values were removed." % (len(df1.index) - len(df2.columns)))
+	#print (df2.head())
+
+	printlog("Reading group file: \"%s\" ..." % (options.group_file))
+	group = pd.read_csv(options.group_file, index_col=0, header=0,names=['Sample_ID', 'Group_ID'])
+	#check if sample IDs are unique
+	if len(group.index) != len(group.index.unique()):
+		print ("Sample IDs are not unique", file = sys.stderr)
+		sys.exit()	
+	group.index = group.index.map(str)
+	printlog("Group file \"%s\" contains %d samples" % (options.group_file, len(group.index)))
+	
+	printlog("Find common sample IDs between group file and data file ...") 
+	common_samples = list(set(group.index) & set(df2.index))
+	used_df = df2.loc[common_samples]
+	(usable_sample, usable_cpg) = used_df.shape
+	printlog("Used CpGs: %d, Used samples: %d" % (usable_cpg, usable_sample))
+
+		
+	printlog("Standarizing values ...")
+	x = used_df.to_numpy()
+	x = StandardScaler().fit_transform(x)
+	
+
+	group_names = group['Group_ID'].unique().tolist()	# a list of unique group names
+	color_names = pick_colors(len(group_names))	# a list of unique colors
+	group_to_col = dict(zip(group_names, color_names))
+	color_list = [group_to_col[g] for g in group['Group_ID']]
+	group['Colors'] = color_list
+	
+	
+	tsne = TSNE(n_components = options.n_components, random_state = 0, perplexity = options.perplexity_value, learning_rate = options.learning_rate, max_iter = options.n_iterations)
+	tsne_components = tsne.fit_transform(x)
+	pc_names = [str(i)+str(j) for i,j in zip(['PC']*options.n_components,range(1,options.n_components+1))]
+	principalDf = pd.DataFrame(data = tsne_components, columns = pc_names, index = used_df.index)
+	principalDf.index.name = 'Sample_ID'
+	
+	finalDf = pd.concat([principalDf, group], axis=1,sort=False, join='inner')
+	finalDf.index.name = 'Sample_ID'
+	
+	printlog("Writing t-SNE results to file: \"%s\" ..." % (options.out_file + '.t-SNE.tsv'))
+	finalDf.to_csv(options.out_file + '.t-SNE.tsv', sep="\t")
+	
+		
+	ROUT = open(options.out_file + '.t-SNE.r','w')
+	
+	print ('pdf(file=\"%s\", width=8, height=8)' % (options.out_file + '.t-SNE.pdf'),file=ROUT)
+	print ('')
+	print ('d = read.table(file=\"%s\", sep="\\t", header=TRUE, comment.char = "", stringsAsFactors=FALSE)' % (options.out_file + '.t-SNE.tsv'), file=ROUT)
+	print ('attach(d)', file=ROUT)
+	
+	if options.plot_alpha:
+		print ('library(scales)', file=ROUT)
+		print ('plot(PC1, PC2, col = alpha(Colors, %f), pch=%d, cex=1.5, main="tSNE 2D map", xlab="tSNE1", ylab="tSNE2")' 
+			% (options.plot_alpha, pch[options.plot_char]), file=ROUT)
+	else:
+		print ('plot(PC1, PC2, col = Colors, pch=%d, cex=1.2, main="tSNE 2D map", xlab="tSNE1", ylab="tSNE2")' 
+			% (pch[options.plot_char]), file=ROUT)
+
+	if options.text_label:
+		print ('text(PC1, PC2, labels=Sample_ID, col = Colors, cex=0.5, pos=1)', file=ROUT)
+	print ('legend("%s", legend=c(%s), col=c(%s), pch=%d,cex=1)' %  (legend_pos[options.legend_location], ','.join(['"' + str(i) + '"' for i in group_names]), ','.join(['"' + str(group_to_col[i]) + '"' for i in group_names]), pch[options.plot_char]), file=ROUT)
+	
+	
+	print ('dev.off()', file=ROUT)
+	ROUT.close()
+	
+	try:
+		subprocess.call("Rscript " + options.out_file + '.t-SNE.r', shell=True)
+	except:
+		print ("Cannot generate pdf file from " + options.out_file + '.t-SNE.r', file=sys.stderr)
+	pass
+
+	
+if __name__=='__main__':
+	main()