def __init__(self, fasta, template_model_filepath, complement_model_filepath):
self.fasta_file = fasta
self.ref_events_dict = {}
self.query_events = []
self.reference_length = 0
self.template_model = EventModel(template_model_filepath)
self.complement_model = EventModel(complement_model_filepath)
self._convert_fast5_to_fasta()
self.event_mapper = EventMapper()
self.samfile = DtwToSam(path_to_template_model_file=os.path.abspath(template_model_filepath),
path_to_complement_model_file=os.path.abspath(complement_model_filepath),
rname=self.fasta_file)
model_kmer_means_complement=process_model_file(model_file_complement)
kmerhashT = process_ref_fasta_raw(fasta_file,model_kmer_means_template)
kmerhashC = process_ref_fasta_raw(fasta_file,model_kmer_means_complement)
seqlengths = get_seq_len(fasta_file)
readprediction=dict()
#SamAlignments=[]
SamOutFile=DtwToSam(path_to_template_model_file=os.path.abspath(model_file_template),path_to_complement_model_file=os.path.abspath(model_file_complement),outfile=args.outsam,rname="EM_079517")
for filename in glob.glob(os.path.join(args.watchdir, '*.fast5')):
print filename
if (args.verbose is True):
print os.path.abspath(filename)
hdf = h5py.File(filename, 'r')
#try:
for read in hdf['Analyses']['EventDetection_000']['Reads']:
events = hdf['Analyses']['EventDetection_000']['Reads'][read]['Events'][()]
event_collection=list()
for event in events:
event_collection.append(float(event[2]))
model_kmer_means_complement=process_model_file(model_file_complement)
kmerhashT = process_ref_fasta_raw(fasta_file,model_kmer_means_template)
kmerhashC = process_ref_fasta_raw(fasta_file,model_kmer_means_complement)
seqlengths = get_seq_len(fasta_file)
readprediction=dict()
#SamAlignments=[]
SamOutFile=DtwToSam(path_to_template_model_file=os.path.abspath(model_file_template),path_to_complement_model_file=os.path.abspath(model_file_complement),outfile=args.outsam,rname="EM_079517")
for filename in glob.glob(os.path.join(args.watchdir, '*.fast5')):
print filename
if (args.verbose is True):
print os.path.abspath(filename)
hdf = h5py.File(filename, 'r')
#try:
for read in hdf['Analyses']['EventDetection_000']['Reads']:
events = hdf['Analyses']['EventDetection_000']['Reads'][read]['Events'][()]
event_collection=list()
for event in events:
print event[2]
#print type event[2]
event_collection.append(event[2])
class Fast52FastaMapper(object):
"""docstring for Fast5ToFastaMapper"""
def __init__(self, fasta, template_model_filepath, complement_model_filepath):
self.fasta_file = fasta
self.ref_events_dict = {}
self.query_events = []
self.reference_length = 0
self.template_model = EventModel(template_model_filepath)
self.complement_model = EventModel(complement_model_filepath)
self._convert_fast5_to_fasta()
self.event_mapper = EventMapper()
self.samfile = DtwToSam(path_to_template_model_file=os.path.abspath(template_model_filepath),
path_to_complement_model_file=os.path.abspath(complement_model_filepath),
rname=self.fasta_file)
def close(self):
self.samfile.close()
def _convert_fast5_to_fasta(self):
## Convert reference to events
self.ref_events_dict = { "template" : {'forward' : [],'reverse' : []},
"complement" : {'forward' : [],'reverse' : []}}
for record in SeqIO.parse(self.fasta_file, 'fasta'):
ref_seq = record.seq
self.reference_length += len(record.seq)
self.ref_events_dict["template"]["forward"].extend(Seq2Events(ref_seq, model = self.template_model).events)
self.ref_events_dict["complement"]["forward"].extend(Seq2Events(ref_seq, model = self.complement_model).events)
self.ref_events_dict["template"]["reverse"].extend(Seq2Events(ref_seq.reverse_complement(), model = self.template_model).events)
self.ref_events_dict["complement"]["reverse"].extend(Seq2Events(ref_seq.reverse_complement(), model = self.complement_model).events)
def map(self, fast5_file):
fast5_file = Fast5File(fast5_file)
template_events, complement_events = self._split_events(fast5_file)
template_start_event = self._calculate_start_event(template_events)
complement_start_event = self._calculate_start_event(complement_events)
template_forward_results = self.event_mapper.map( self.ref_events_dict["template"]["forward"], template_events[template_start_event:] )
template_reverse_results = self.event_mapper.map( self.ref_events_dict["template"]["reverse"], template_events[template_start_event:] )
self._convert_path_to_template_postitions(template_forward_results.path, template_start_event)
self._convert_path_to_template_postitions(template_reverse_results.path, template_start_event, forward = False)
if complement_events:
complement_forward_results = sself.event_mapper.map( self.ref_events_dict["complement"]["forward"], complement_events[complement_start_event:] )
complement_reverse_results = sself.event_mapper.map( self.ref_events_dict["complement"]["reverse"], complement_events[complement_start_event:] )
self._convert_path_to_template_postitions(complement_forward_results.path, template_start_event)
self._convert_path_to_template_postitions(complement_reverse_results.path, template_start_event, forward = False)
def _convert_path_to_template_postitions(self, path, start_event, forward = True):
print path
f,t = path
for (f,t) in zip(f,t):
read_position = f + start_event
if not forward:
ref_position = self.reference_length - t
# print "reference length", self.reference_length
else:
ref_position = t
# print [read_position,ref_position]
def _calculate_start_event(self, events):
prev1 = 0
prev2 = 0
counter = 0
start_event = 0
for event in events:
if (prev1 - event.mean > 40 ) or (prev2 - event.mean > 40):
start_event=counter+1
break
prev2 = prev1
prev1 = event.mean
counter +=1
return start_event
def _split_events(self, fast5_file):
events = fast5_file.get_pre_basecalled_events()
if fast5_file.read_metadata[0].get('has_hairpin',False):
template_events = events[:self.fast5_file.read_metadata['hairpin_event_index'] ]
complement_events = events[self.fast5_file.read_metadata['hairpin_event_index']:]
else:
template_events = events
complement_events = []
return template_events, complement_events
