sync

medaka/medaka.py

@@ -10,7 +10,7 @@ def main():
 	args = parse_arguments()
 	mini_align(args)
 	medaka_consensus(args)
-
+	medaka_stich(args)
 
 def parse_arguments():
 	# Initialize parser
@@ -23,7 +23,7 @@ def parse_arguments():
 	parser.add_argument("-m", help = "fill MD tag")
 	parser.add_argument("-p", help = "output file prefix (default: reads)",nargs='?', default="reads", type=str)
 	parser.add_argument("-c", "--contig-chunks", help = "Number of contigs per medaka_consensus step", nargs='?',default=100, type=int)
-	parser.add_argument("--model", help = "Model definition, default is equivalent to r94.", default="r94")
+	parser.add_argument("--model", help = "Model definition, default is equivalent to r941_min_fast_g303.", default="r941_min_fast_g303")
 	# Adding optional argument for ....
 	# ...
 
@@ -42,9 +42,10 @@ def parse_arguments():
 #     -p calls_to_draft.bam -t <threads>
 
 def mini_align(args):
-	
-	if os.path.exists(args.reference + ".mmi"):
-		print("Index file is already present")
+	# Checks for mmi and bam file?
+	if os.path.exists(args.reference + ".mmi") and os.path.exists(args.p + ".bam"):
+		print("Index and bam file are already present")
+		return
 	print(args)
 	arguments = ["mini_align", "-i", args.input, "-r", args.reference, "-m", "-p", args.p, "-t", args.threads]
 	command = ' '.join(map(str,arguments))
@@ -94,12 +95,12 @@ def medaka_consensus(args):
 	reader.close()
 
 	# Run this in parallel for each chunk by number of threads / 2
-	commands = set()
+	commands = []
 	for chunk in chunks:
 		arguments = ["medaka", "consensus", args.p + ".bam", "results/chunk_" + str(chunk), "--model", args.model, "--threads", "2", "--region", ' '.join(chunks[chunk])]
 		arguments = ' '.join(map(str,arguments))
-		commands.add(arguments)
-
+		commands.append(arguments)
+	
 	# Multi thread section
 	from functools import partial
 	from multiprocessing.dummy import Pool
@@ -110,9 +111,9 @@ def medaka_consensus(args):
 	pool = Pool(jobs) # two concurrent commands at a time
 	for i, returncode in enumerate(pool.imap(partial(call, shell=True), commands)):
 		if returncode != 0:
-			print("%d command failed: %d" % (i, returncode))
+			print("%d command failed with exit code %d and command %s" % (i, returncode, commands[i]))
 		else:
-			print("%d command finished: %d" % (i, returncode))
+			print("%d command finished: %d %s" % (i, returncode, commands[i]))
 
 
 ## SINGLE RUN