From 2742e59d5d4d79b99f67ba70933667c3cff6f229 Mon Sep 17 00:00:00 2001
From: govinda-kamath <govinda.kamath@gmail.com>
Date: Thu, 19 Jan 2017 08:41:55 -0800
Subject: [PATCH] Added missing file

---
 Zeisel_pipeline/subsample_bams.py | 111 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 111 insertions(+)
 create mode 100644 Zeisel_pipeline/subsample_bams.py

diff --git a/Zeisel_pipeline/subsample_bams.py b/Zeisel_pipeline/subsample_bams.py
new file mode 100644
index 0000000..9ef0f6b
--- /dev/null
+++ b/Zeisel_pipeline/subsample_bams.py
@@ -0,0 +1,111 @@
+import sys
+import pysam
+import numpy as np
+import os
+import itertools
+import pickle
+import multiprocessing as mp
+import getopt
+
+try:
+    opts, args = getopt.getopt(sys.argv[1:],"s:n:b:o:",["subsampled-read-dir=","njobs=","base-bam-dir=","out-bam-dir="])
+except getopt.GetoptError:
+    print ("getopterrror")
+    print ('usage is : \n python subsample_bams.py -s subsampled-read-dir -b base-bam-dir -o out-bam-dir [-n number-of-processes-to-use]')
+    sys.exit(1)
+    
+
+num_proc=1
+bamfile_base_dir=''
+output_base_dir=''
+sub_reads_base_dir = ''
+
+for opt,arg in opts:
+    if opt in ("-s", "--subsampled-read-dir"):
+        sub_reads_base_dir=arg
+    elif opt in ("-n","--njobs"):
+        num_proc=int(arg)
+    elif opt in ("-b","--base-bam-dir"):
+        bamfile_base_dir=arg
+    elif opt in ("-o","--out-bam-dir"):
+        output_base_dir=arg
+
+if (not bamfile_base_dir) or (not output_base_dir) or (not sub_reads_base_dir):
+    print ('usage is : \n python subsample_bams.py -s subsampled-read-dir -b base-bam-dir -o out-bam-dir [-n number-of-processes-to-use]')
+    sys.exit(1)
+# In[5]:
+
+def subsample_bam(celltuple):
+    cellname=celltuple[0]
+    bamfile_base_dir=celltuple[1]
+    output_base_dir=celltuple[2]
+    sub_reads_base_dir = celltuple[3]
+    bamfile_path=bamfile_base_dir+cellname+'/'+'hits.bam'
+    
+    #bamfile_path='/data/SS_RNA_seq/Zeisel/samtools_experiments/little.bam'
+    print cellname
+#    print '\n'+'loading bamfile...'
+    ## INPUT 1 -- original bamfile
+    bamfile = pysam.AlignmentFile(bamfile_path, "rb")
+#    print 'sorting bam...'
+    read_dict={}
+    for read in bamfile:
+        if not read.is_unmapped:
+            read_dict.setdefault(int(read.query_name.split('.')[1]),[])
+            read_dict[int(read.query_name.split('.')[1])].append(read)
+        
+    sampling_suffix=['10','5','1','_point5', '_point1']
+
+    
+    for suffix in sampling_suffix:
+#        print 'in '+suffix+'...'
+        
+        out_dir=output_base_dir+suffix+'/'
+        os.system('mkdir -p '+ out_dir)
+        out_cell_dir=out_dir+cellname+'/'
+        os.system('mkdir -p '+ out_cell_dir)
+        outfile_path=out_cell_dir+'hits.bam'
+        
+        sub_read_dir=sub_reads_base_dir+suffix+'/'
+        sub_reads_path=sub_read_dir+cellname+'.rnum'
+        umi_path=sub_read_dir+cellname+'.umi'
+        sub_reads = sorted(np.loadtxt(sub_reads_path,dtype=int))
+        umi=np.loadtxt(umi_path,dtype=str,usecols=(0,))
+        
+#        print 'writing header...'
+        ## OUTPUT -- new bam file with the same header
+        outfile = pysam.AlignmentFile(outfile_path, "wb", header=bamfile.header)
+        
+        mapped_UMI=[]
+        unmapped_UMI=[]
+#        print 'subsampling reads...'
+        for index in xrange(len(sub_reads)):
+            read_id=sub_reads[index]
+            try:
+                for read in read_dict[read_id]: 
+                    outfile.write(read)
+                mapped_UMI.append(umi[index])
+            except KeyError:
+                unmapped_UMI.append(umi[index])
+                
+        outfile.close()
+        
+#        print 'writing UMIs ...'
+        
+        np.savetxt(out_cell_dir+'mapped.umi',np.array(mapped_UMI),delimiter="\n", fmt="%s")
+        np.savetxt(out_cell_dir+'unmapped.umi',np.array(unmapped_UMI),delimiter="\n", fmt="%s")
+        
+#         with open(out_cell_dir+'mapped.umi', 'wb') as f:
+#             pickle.dump(mapped_UMI,f,pickle.HIGHEST_PROTOCOL)
+#         with open(out_cell_dir+'unmapped.umi', 'wb') as f:
+#             pickle.dump(unmapped_UMI,f,pickle.HIGHEST_PROTOCOL)
+#        print 'DONE.'
+
+
+# In[6]:
+
+cellnames=sorted([x for x in os.listdir(bamfile_base_dir) if x.startswith('SRR')])
+
+celltuples=itertools.product(cellnames,[bamfile_base_dir],[output_base_dir],[sub_reads_base_dir])
+pool=mp.Pool(processes=num_proc)
+pool.map(subsample_bam,celltuples)