def map_libraries(self):
"""Map to arbitrary rna sequence libraries"""
out = self.output
libraries = self.libraries
if libraries == '' or len(libraries) == 0:
print('no libraries to map to')
return
#map to provided libraries
print('mapping to these libraries: %s' % libraries)
res, counts = base.map_rnas(self.files,
libraries,
self.temp_path,
aligner=self.aligner,
samplelabels=self.labels,
params=self.aligner_params)
if res is None:
print('empty data returned. did alignments run?')
return
print('results saved to rna_counts.csv')
res.to_csv(os.path.join(out, 'rna_found.csv'), index=False)
counts.to_csv(os.path.join(out, 'rna_counts.csv'), index=False)
plot_results(res, out)
return
def map_libraries(self):
"""Map to arbitrary rna sequence libraries"""
out = self.output
libraries = self.libraries
if libraries == '' or len(libraries) == 0:
print ('no libraries to map to')
return
#map to provided libraries
print ('mapping to these libraries: %s' %libraries)
res, counts = base.map_rnas(self.files, libraries, self.temp_path,
aligner=self.aligner,
samplelabels=self.labels,
params=self.aligner_params)
if res is None:
print ('empty data returned. did alignments run?')
return
print ('results saved to rna_counts.csv')
res.to_csv( os.path.join(out, 'rna_found.csv'),index=False)
counts.to_csv( os.path.join(out, 'rna_counts.csv'), index=False )
plot_results(res, out)
return
def map_mirnas(self):
"""Map miRNAs using mirbase with isomir counts and do novel prediction
if a reference genome and index is provided"""
out = self.output
libraries = self.libraries
temp = self.temp_path
ref_name = self.ref_name
mat_name = 'mirbase-%s' %self.species
self.aligner_params[mat_name] = self.mirna_params
novel.VERBOSE = self.verbose
if self.check_index(ref_name) == False:
print ('no index for reference genome')
ref_name = ''
print ('mapping miRNAs..')
res, counts = base.map_mirbase(self.files, outpath=temp, indexes=libraries,
species=self.species, ref_genome=ref_name,
pad5=self.pad5, pad3=self.pad3, aligner=self.aligner,
samplelabels=self.labels,
params=self.aligner_params,
verbose=self.verbose)
#seperate out mature counts and save
matcounts = counts[counts.ref==mat_name]
res.to_csv( os.path.join(out, 'results.csv'),index=False )
res = res[res.ref!=ref_name]
matcounts.to_csv( os.path.join(out, 'mirbase_mature_counts.csv'), index=False,
float_format='%.1f' )
counts.to_csv( os.path.join(out, 'all_counts.csv'), index=False, float_format='%.1f')
plot_results(res, out)
#isomir counting
print ()
print ('counting isomirs..')
iso, isocounts = base.map_isomirs(self.files, temp, self.species,
samplelabels=self.labels)
isocounts.to_csv( os.path.join(out, 'isomir_counts.csv'), index=False, float_format='%.1f')
#novel prediction
if self.ref_fasta == '' or not os.path.exists(self.ref_fasta):
print ('no reference genome file, skipping novel mirna step')
elif ref_name == None or ref_name == '':
print ('no index for ref genome, required for novel mirna step')
elif check_viennarna() == False:
print ('Vienna RNA package not installed')
print ('see https://www.tbi.univie.ac.at/RNA/')
else:
print ()
print ('predicting novel mirnas..')
start = time.time()
#change map_rnas so it can use remaining files from previous run....?
allreads = utils.combine_aligned_reads(temp, idx=ref_name)
new,cl = novel.find_mirnas(allreads, self.ref_fasta, species=self.species,
score_cutoff=float(self.score_cutoff),
read_cutoff=int(self.read_cutoff),
cpus=self.cpus)
if new is None or len(new) == 0:
print ('could not find any novel mirnas at this score cutoff')
return
if self.strict == True:
new = new[new.mature_check=='ok']
print ('filtered %s' %len(new))
new.to_csv(os.path.join(out,'novel_mirna.csv'), index=False)
#pad mature novel and write to fasta for counting
novpad = base.get_mature_padded(new, idkey='mature_id', seqkey='mature')
novpad = novpad.drop_duplicates('name')
utils.dataframe_to_fasta(novpad,os.path.join(out,'novel.fa'),
seqkey='sequence', idkey='name')
novel.create_report(new, cl, self.species, outfile=os.path.join(out, 'novel.html'))
#now count novel mirnas for all samples
build_indexes(os.path.join(out,'novel.fa'), self.index_path)
r,nc = base.map_rnas(self.files, ['novel'], self.temp_path,
aligner=self.aligner,
samplelabels=self.labels)
nc.to_csv( os.path.join(out, 'novel_mirna_counts.csv'), index=False )
end = round(time.time()-start,1)
print ('took %s seconds' %str(end))
return
def map_mirnas(self):
"""Map miRNAs using mirbase with isomir counts and do novel prediction
if a reference genome and index is provided"""
out = self.output
libraries = self.libraries
temp = self.temp_path
ref_name = self.ref_name
mat_name = 'mirbase-%s' %self.species
self.aligner_params[mat_name] = self.mirna_params
novel.VERBOSE = self.verbose
if self.check_index(ref_name) == False:
print ('no index for reference genome')
ref_name = ''
print ('mapping miRNAs..')
res, counts = base.map_mirbase(self.files, outpath=temp, indexes=libraries,
species=self.species, ref_genome=ref_name,
pad5=self.pad5, pad3=self.pad3, aligner=self.aligner,
samplelabels=self.labels,
params=self.aligner_params,
verbose=self.verbose)
self.results = res
#seperate out mature counts and save
matcounts = counts[counts.ref==mat_name]
res.to_csv( os.path.join(out, 'results.csv'),index=False )
res = res[res.ref!=ref_name]
matcounts.to_csv( os.path.join(out, 'mirbase_mature_counts.csv'), index=False,
float_format='%.1f' )
counts.to_csv( os.path.join(out, 'all_counts.csv'), index=False, float_format='%.1f')
#get fractions per sample and plot results
c = base.pivot_count_data(res, idxcols=['name','ref'])
self.samples = s = base.get_fractions_mapped(res)
print (s)
plot_results(s, c, out)
#isomir counting
print ()
print ('counting isomirs..')
iso, isocounts = base.map_isomirs(self.files, temp, self.species,
samplelabels=self.labels)
if isocounts is not None:
isocounts.to_csv( os.path.join(out, 'isomir_counts.csv'),
index=False, float_format='%.1f')
else:
print ('no isomirs could be counted')
#novel prediction
#train classifier first if not present
novel.create_classifier()
if self.ref_fasta == '' or not os.path.exists(self.ref_fasta):
print ('no reference genome file, skipping novel mirna step')
elif ref_name == None or ref_name == '':
print ('no index for ref genome, required for novel mirna step')
elif check_viennarna() == False:
print ('Vienna RNA package not installed')
print ('see https://www.tbi.univie.ac.at/RNA/')
else:
print ()
print ('predicting novel mirnas..')
start = time.time()
#change map_rnas so it can use remaining files from previous run....?
allreads = utils.combine_aligned_reads(temp, idx=ref_name)
new,cl = novel.find_mirnas(allreads, self.ref_fasta, species=self.species,
score_cutoff=float(self.score_cutoff),
read_cutoff=int(self.read_cutoff),
cpus=self.cpus)
if new is None or len(new) == 0:
print ('Could not find any novel mirnas.')
print ('There may not be sufficient aligned reads or the score cutoff is too high.\n')
return
if self.strict == True:
new = new[new.mature_check=='ok']
print ('filtered %s' %len(new))
new.to_csv(os.path.join(out,'novel_mirna.csv'), index=False)
#pad mature novel and write to fasta for counting
novpad = base.get_mature_padded(new, idkey='mature_id', seqkey='mature')
novpad = novpad.drop_duplicates('name')
utils.dataframe_to_fasta(novpad,os.path.join(out,'novel.fa'),
seqkey='sequence', idkey='name')
novel.create_report(new, cl, self.species, outfile=os.path.join(out, 'novel.html'))
#now count novel mirnas for all samples
build_indexes(os.path.join(out,'novel.fa'), self.index_path)
r,nc = base.map_rnas(self.files, ['novel'], self.temp_path,
aligner=self.aligner,
samplelabels=self.labels)
nc.to_csv( os.path.join(out, 'novel_mirna_counts.csv'), index=False )
end = round(time.time()-start,1)
print ('took %s seconds' %str(end))
return