upload scripts

Author: Aris-K

annotation.py

@@ -0,0 +1,76 @@
+import numpy as np
+import sys
+import dihmm_ext
+import bed_writer
+import os
+from os import listdir,mkdir
+from os.path import isfile,join,exists
+from optparse import OptionParser
+import pandas as pd
+
+def printlist(path,value,sep):
+    # "print the list to text file"
+    fid=open(path,'w')
+    nline=[]
+    for listvalue in value:
+        if isinstance(listvalue[0],float):
+            nline=sep.join(str(i) for i in listvalue)
+        elif isinstance(listvalue[0],int):
+            nline=sep.join(str(i) for i in listvalue)
+        elif isinstance(listvalue[0],str):
+            nline=sep.join(listvlaue)
+        fid.write("%s\n" % nline)
+    fid.close()
+
+def main(argv):
+    parser = OptionParser()
+    parser.add_option("-t", "--table", action="store", type="string", dest="table", metavar="<file>", help="the file for currently annotated samples")
+    parser.add_option("-n", "--name", action="store", type="string", dest="name", metavar="<file>", help="the name of the annotation pair to work on")
+    parser.add_option("-i", "--indata", action="store", type="string", dest="indata", metavar="<file>", help="the path for prepared signal matrix in the target cell type")
+    parser.add_option("-m", "--model", action="store", type="string", dest="model", metavar="<file>", help="the path for trained model")
+    parser.add_option("-o", "--outfolder", action="store", type="string", dest="outfolder", metavar="<file>", help="the path for output folder to store the annotated chroms")
+    
+    (opt, args) = parser.parse_args(argv)
+    if len(argv) < 8:
+        parser.print_help()
+        sys.exit(1)
+    
+    domain_size = 20
+    path1=opt.indata
+    file1=[path1+f for f in listdir(path1)]
+    allfile=file1
+
+    cdf = pd.read_csv(opt.table, sep='\t', header=None)
+    currentAnnotations = list(cdf[0])
+    name = opt.name
+    if name not in currentAnnotations:
+        print("working on "+name+"...")
+        os.mkdir(opt.outfolder)
+        x=dihmm_ext.load_model(opt.model+'/',domain_size)
+        p=x.emission_probabilities
+        bt=x.bin_transition_probabilities
+        dt=x.domain_transition_probabilities
+        nb=x.n_bin_states
+        nd=x.n_domain_states
+
+        output=opt.outfolder+'/'
+        for file in allfile:
+            tmp=file.split('/')[-1].split('_')
+            cellline=tmp[0]
+
+            chrom=tmp[1]
+            noutput=output+cellline+'/'
+            if not os.path.exists(noutput):
+                os.mkdir(noutput)
+            a=dihmm_ext.annotate(x,[file])
+            b=bed_writer.BedWriter(a[0],x)
+            b.write_bed_files(noutput,cellline,chrom)
+            anno=a[0].annotations
+            bsd=a[0].bin_state_distributions
+            dsd=a[0].domain_state_distributions
+            printlist(noutput+chrom+'_bin_domain_states.txt',anno.tolist(),'\t')
+    else:
+        print(opt.name+" has already exsited...skip...")
+        
+if __name__ == "__main__":
+    main(sys.argv)

train.py

@@ -0,0 +1,26 @@
+import numpy as np
+import dihmm_ext
+import bed_writer
+import os
+from os import listdir,mkdir
+from os.path import isfile,join,exists
+
+n_bin_states = 30
+n_domain_states = 30
+domain_size = 20
+tolerance = 1e-6
+max_iter = 500
+
+
+output='/data/sysdir/home/yk890/diHMM/diHMM/chr17_3cells/model'
+path='/data/sysdir/home/yk890/diHMM/diHMM/chr17_3cells/data'
+
+tpath=path+'/'
+trainfile=[tpath+f for f in listdir(tpath)]
+x=dihmm_ext.run_dihmm(n_bin_states, n_domain_states, domain_size, max_iter, tolerance, trainfile)
+toutput=output+''
+if not os.path.exists(toutput):
+    os.mkdir(toutput)
+dihmm_ext.save_model(x,toutput)
+
+