def run(self):
extractions = [ ]
for item in self.genes.split(','):
extraction = item.split('/')
assert len(extraction) == 4
extractions.append(extraction)
rename = { }
if self.rename:
for item in self.rename.split(','):
old,new = item.split('=')
rename[old] = new
work = self.get_workspace()
with workspace.tempspace() as temp:
items = list(annotation.read_annotations(self.annotation))
for item in items:
item.seqid = rename.get(item.seqid, item.seqid)
annotation.write_gff3(temp/'temp.gff', get_genes(items, extractions, self.log))
del items
with open(temp/'temp.fa','wb') as f:
for name,seq in io.read_sequences(self.genome):
name = name.split()[0]
name = rename.get(name,name)
io.write_fasta(f, name, seq)
reference_directory.Make_tt_reference(
self.output_dir,
filenames = [ temp/'temp.fa', temp/'temp.gff' ] + self.extra,
index = self.index, shrimp = self.shrimp,
bowtie = self.bowtie, star = self.star
).run()
def run(self):
extractions = [ ]
for item in self.genes.split(','):
extraction = item.split('/')
assert len(extraction) == 4
extractions.append(extraction)
rename = { }
if self.rename:
for item in self.rename.split(','):
old,new = item.split('=')
rename[old] = new
work = self.get_workspace()
with workspace.tempspace() as temp:
items = list(annotation.read_annotations(self.annotation))
for item in items:
item.seqid = rename.get(item.seqid, item.seqid)
annotation.write_gff3(temp/'temp.gff', get_genes(items, extractions, self.log))
del items
with open(temp/'temp.fa','wb') as f:
for name,seq in io.read_sequences(self.genome):
name = name.split()[0]
name = rename.get(name,name)
io.write_fasta(f, name, seq)
reference_directory.Make_tt_reference(
self.output_dir,
filenames = [ temp/'temp.fa', temp/'temp.gff' ],
index = self.index,
).run()
def run(self):
assert self.release
assert self.species
assert self.assembly
assert self.dna
extractions = [ ]
for item in self.genes.split(','):
extraction = item.split('/')
assert len(extraction) == 4
extractions.append(extraction)
rename = { }
if self.rename:
for item in self.rename.split(','):
old,new = item.split('=')
rename[old] = new
work = self.get_workspace()
ensembl = workspace.Workspace(work/'ensembl')
genome_filename = self.species+"."+self.assembly+"."+self.dna+".fa.gz"
genome_url = "rsync://ftp.ensembl.org/ensembl/pub/release-"+self.release+"/fasta/"+self.species.lower()+"/dna/"+genome_filename
gff_filename = self.species+"."+self.assembly+"."+self.release+".gff3.gz"
gff_url = "rsync://ftp.ensembl.org/ensembl/pub/release-"+self.release+"/gff3/"+self.species.lower()+"/"+gff_filename
if self.download:
self.log.log("Fetching "+genome_url+"\n")
io.execute(['rsync','-aP',genome_url, ensembl/genome_filename])
self.log.log("Fetching "+gff_url+"\n")
io.execute(['rsync','-aP',gff_url, ensembl/gff_filename])
with workspace.tempspace() as temp:
items = list(annotation.read_annotations(ensembl/gff_filename))
for item in items:
item.seqid = rename.get(item.seqid, item.seqid)
annotation.write_gff3(temp/'temp.gff', get_genes(items, extractions, self.log))
del items
with open(temp/'temp.fa','wb') as f:
for name,seq in io.read_sequences(ensembl/genome_filename):
name = name.split()[0]
name = rename.get(name,name)
io.write_fasta(f, name, seq)
reference_directory.Make_tt_reference(
self.output_dir,
filenames = [ temp/'temp.fa', temp/'temp.gff' ],
index = self.index,
).run()
def run(self):
items = list(annotation.read_annotations(self.parent))
annotation.link_up_annotations(items)
for item in items:
assert len(item.parents) <= 1
genes = [ item for item in items if item.type == "gene" ]
downstrand_genes = [ _extend(_three_prime(item), self.extension) for item in genes ]
exons = [ item for item in items if item.type == "exon" ]
utrs = [ _extend(item, self.extension) for item in items if item.type == "three_prime_utr" ]
gene_index = span_index.index_annotations(genes)
downstrand_gene_index = span_index.index_annotations(downstrand_genes)
exon_index = span_index.index_annotations(exons)
utr_index = span_index.index_annotations(utrs)
peaks = list(annotation.read_annotations(self.child))
for peak in peaks:
# Query is final base in genome before poly(A) starts
query = peak.three_prime().shifted(-1,0)
hit_to = "3'UTR"
hits = [ item.parents[0].parents[0] for item in
utr_index.get(query, True) ]
if not hits:
hit_to = "Exon"
hits = [ item.parents[0].parents[0] for item in
exon_index.get(query, True) ]
# For non-coding RNAs, which don't have a 3' UTR
if not hits:
hit_to = "Downstrand"
hits = downstrand_gene_index.get(query, True)
if not hits:
hit_to = "Intron"
hits = gene_index.get(query, True)
antisense_hits = gene_index.get(query.reversed(), True)
if not hits:
hit_to = "Antisense"
hits = antisense_hits
if hits:
peak.attr["Parent"] = join_descriptions([ item.get_id() for item in hits ], ",")
peak.attr["Relation"] = hit_to
peak.attr["Name"] = join_descriptions([ item.attr.get("Name","") for item in hits ])
peak.attr["Product"] = hit_to + " " + join_descriptions([ item.attr.get("Product","") for item in hits ])
peak.attr["Biotype"] = join_descriptions([ item.attr.get("Biotype","") for item in hits ])
if antisense_hits:
peak.attr["Antisense_parent"] = join_descriptions([ item.get_id() for item in antisense_hits ], ",")
peak.attr["Antisense_name"] = join_descriptions([ item.attr.get("Name","") for item in antisense_hits ])
peak.attr["Antisense_product"] = "Antisense " + join_descriptions([ item.attr.get("Product","") for item in antisense_hits ])
peak.attr["Antisense_biotype"] = join_descriptions([ item.attr.get("Biotype","") for item in antisense_hits])
annotation.write_gff3(self.prefix+"-parent.gff", genes) #Hmm
annotation.write_gff3(self.prefix+"-child.gff", peaks)
def run(self):
seqs = env.load_ref(self.reference).seqs
result = []
errors = []
with open(self.csv_file, "rU") as f:
reader = csv.reader(f)
headings = reader.next()
headings = [item.lower() for item in headings]
assert "id" in headings
assert "primer" in headings
id_col = headings.index("id")
primer_col = headings.index("primer")
for row in reader:
if len(row) == 0 or (not row[id_col].strip()
and not row[primer_col].strip()):
continue
id = row[id_col].strip()
assert " " not in id, "ID contains space: " + id
primer = row[primer_col].strip().upper()
assert len(primer) > self.skip, "Primer too short: " + id
assert [char in "ACGT"
for char in primer], "Primer not ACGT: " + id
primer = primer[self.skip:]
rprimer = bio.reverse_complement(primer)
hits = []
for seq_name in seqs:
for match in re.finditer(primer, seqs[seq_name],
re.IGNORECASE):
hits.append((seq_name, 1, match.start(),
match.start() + self.length))
for match in re.finditer(rprimer, seqs[seq_name],
re.IGNORECASE):
hits.append((seq_name, -1, match.end() - self.length,
match.end()))
if len(hits) > 100:
raise config.Error("Many many hits for " + id + ".")
if not hits:
errors.append("No hits for " + id + ".")
continue
if len(hits) > 1:
self.log.log("Warning: %d hits for %s.\n" %
(len(hits), id))
for i, hit in enumerate(hits):
hit_name = id
if len(hits) > 1:
hit_name += "-%dof%d" % (i + 1, len(hits))
result.append(
annotation.Annotation(seqid=hit[0],
source="tail-tools",
type="region",
start=hit[2],
end=hit[3],
strand=hit[1],
attr=dict(ID=hit_name,
Primer=primer)))
if errors:
raise config.Error("\n".join(errors))
annotation.write_gff3(self.prefix + ".gff", result)
def run(self):
work = self.get_workspace()
work.update_param(remove=['tail_tools_reference_version'])
nesoni.Make_reference(
self.output_dir,
filenames = self.filenames,
snpeff = False,
cs = 'ifavailable' if self.index else False,
ls = False,
bowtie = 'ifavailable' if self.index else False,
).run()
annotations = list(annotation.read_annotations(work/'reference.gff'))
annotation.link_up_annotations(annotations)
exon_index = span_index.index_annotations([
item for item in annotations if item.type == "exon"
])
mrna_end_index = span_index.index_annotations([
item.three_prime() for item in annotations if item.type == "mRNA"
])
mrna_utrs = [ ]
gene_utrs = [ ]
for gene in annotations:
if gene.type != 'gene': continue
mrnas = [ item for item in gene.children if item.type == 'mRNA' ]
assert mrnas, "Gene without any mRNAs: "+gene.get_id()
gene.attr['color'] = '#880088'
gene.start = min(item.start for item in mrnas)
gene.end = max(item.end for item in mrnas)
gene.attr["max_extension"] = str(_max_extension(gene, exon_index, mrna_end_index))
gene_utr_5primes = [ ]
for mrna in mrnas:
assert mrna.strand == gene.strand, mrna
assert mrna.seqid == gene.seqid, mrna
mrna.attr["max_extension"] = str(_max_extension(mrna, exon_index, mrna_end_index))
cdss = [ item for item in mrna.children if item.type == 'CDS' ]
exons = [ item for item in mrna.children if item.type == 'exon' ]
if not exons: continue
#link up annotations sorts children, so final is really final
for item in exons[:-1]:
item.attr["max_extension"] = "0"
exons[-1].attr["max_extension"] = mrna.attr["max_extension"]
if not cdss: continue
mrna_utr_5primes = [ ]
if gene.strand >= 0:
cds_3prime = max(item.end for item in cdss)
for item in exons:
if item.end >= cds_3prime:
mrna_utr_5primes.append(max(item.start,cds_3prime))
else:
cds_3prime = min(item.start for item in cdss)
for item in exons:
if item.start <= cds_3prime:
mrna_utr_5primes.append(min(item.end,cds_3prime))
if mrna.strand >= 0:
utr_start = min(mrna_utr_5primes) if mrna_utr_5primes else mrna.end
utr_end = max(utr_start+1,mrna.end)
gene_utr_5primes.append(utr_start)
else:
utr_end = max(mrna_utr_5primes) if mrna_utr_5primes else mrna.start
utr_start = min(mrna.start,utr_end-1)
gene_utr_5primes.append(utr_end)
attr = mrna.attr.copy()
attr['Parent'] = attr['ID']
attr['ID'] = attr['ID']+'-3UTR'
attr['color'] = '#008888'
utr = annotation.Annotation(
source = 'tt',
type = 'three_prime_utr',
seqid = mrna.seqid,
strand = mrna.strand,
start = utr_start,
end = utr_end,
attr = attr,
)
max_ext = _max_extension(utr, exon_index, mrna_end_index)
utr.attr["max_extension"] = str(max_ext)
#Only include if there is an annotated 3' UTR or end is not in the middle of some other isoform's exon
if utr_end-utr_start+max_ext > 1:
mrna_utrs.append(utr)
if gene.strand >= 0:
utr_start = max(gene_utr_5primes) if gene_utr_5primes else gene.end
utr_end = max(utr_start+1,gene.end)
else:
utr_end = min(gene_utr_5primes) if gene_utr_5primes else gene.start
utr_start = min(gene.start,utr_end-1)
attr = gene.attr.copy()
attr['Parent'] = attr['ID']
attr['ID'] = attr['ID']+'-3UTR'
attr['color'] = '#008888'
utr = annotation.Annotation(
source = 'tt',
type = 'three_prime_utr',
seqid = gene.seqid,
strand = gene.strand,
start = utr_start,
end = utr_end,
attr = attr,
)
utr.attr["max_extension"] = str(_max_extension(utr, exon_index, mrna_end_index))
gene_utrs.append(utr)
annotation.write_gff3(work/'reference.gff', annotations + mrna_utrs)
annotation.write_gff3(work/'utr.gff', gene_utrs)
work.update_param(tail_tools_reference_version=work.VERSION)
def run(self):
seqs = env.load_ref(self.reference).seqs
result = [ ]
errors = [ ]
with open(self.csv_file, "rU") as f:
reader = csv.reader(f)
headings = reader.next()
headings = [ item.lower() for item in headings ]
assert "id" in headings
assert "primer" in headings
id_col = headings.index("id")
primer_col = headings.index("primer")
for row in reader:
if len(row) == 0 or (not row[id_col].strip() and not row[primer_col].strip()):
continue
id = row[id_col].strip()
assert " " not in id, "ID contains space: "+id
primer = row[primer_col].strip().upper()
assert len(primer) > self.skip, "Primer too short: "+id
assert [ char in "ACGT" for char in primer ], "Primer not ACGT: "+id
primer = primer[self.skip:]
rprimer = bio.reverse_complement(primer)
hits = [ ]
for seq_name in seqs:
for match in re.finditer(
primer, seqs[seq_name], re.IGNORECASE):
hits.append( (seq_name, 1, match.start(), match.start()+self.length) )
for match in re.finditer(
rprimer, seqs[seq_name], re.IGNORECASE):
hits.append( (seq_name, -1, match.end()-self.length, match.end()) )
if len(hits) > 100:
raise config.Error("Many many hits for "+id+".")
if not hits:
errors.append("No hits for "+id+".")
continue
if len(hits) > 1:
self.log.log("Warning: %d hits for %s.\n" % (len(hits),id))
for i, hit in enumerate(hits):
hit_name = id
if len(hits) > 1:
hit_name += "-%dof%d" % (i+1,len(hits))
result.append(annotation.Annotation(
seqid = hit[0],
source = "tail-tools",
type = "region",
start = hit[2],
end = hit[3],
strand = hit[1],
attr = dict(
ID=hit_name,
Primer=primer
)
))
if errors:
raise config.Error("\n".join(errors))
annotation.write_gff3(self.prefix+".gff", result)
def run(self):
work = self.get_workspace()
work.update_param(remove=['tail_tools_reference_version'])
nesoni.Make_reference(
self.output_dir,
filenames=self.filenames,
snpeff=False,
cs='ifavailable' if self.index else False,
ls=False,
bowtie='ifavailable' if self.index else False,
).run()
annotations = list(annotation.read_annotations(work / 'reference.gff'))
annotation.link_up_annotations(annotations)
exon_index = span_index.index_annotations(
[item for item in annotations if item.type == "exon"])
mrna_end_index = span_index.index_annotations([
item.three_prime() for item in annotations if item.type == "mRNA"
])
mrna_utrs = []
gene_utrs = []
for gene in annotations:
if gene.type != 'gene': continue
mrnas = [item for item in gene.children if item.type == 'mRNA']
assert mrnas, "Gene without any mRNAs: " + gene.get_id()
gene.attr['color'] = '#880088'
gene.start = min(item.start for item in mrnas)
gene.end = max(item.end for item in mrnas)
gene.attr["max_extension"] = str(
_max_extension(gene, exon_index, mrna_end_index))
gene_utr_5primes = []
for mrna in mrnas:
assert mrna.strand == gene.strand, mrna
assert mrna.seqid == gene.seqid, mrna
mrna.attr["max_extension"] = str(
_max_extension(mrna, exon_index, mrna_end_index))
cdss = [item for item in mrna.children if item.type == 'CDS']
exons = [item for item in mrna.children if item.type == 'exon']
if not exons: continue
#link up annotations sorts children, so final is really final
for item in exons[:-1]:
item.attr["max_extension"] = "0"
exons[-1].attr["max_extension"] = mrna.attr["max_extension"]
if not cdss: continue
mrna_utr_5primes = []
if gene.strand >= 0:
cds_3prime = max(item.end for item in cdss)
for item in exons:
if item.end >= cds_3prime:
mrna_utr_5primes.append(max(
item.start, cds_3prime))
else:
cds_3prime = min(item.start for item in cdss)
for item in exons:
if item.start <= cds_3prime:
mrna_utr_5primes.append(min(item.end, cds_3prime))
if mrna.strand >= 0:
utr_start = min(
mrna_utr_5primes) if mrna_utr_5primes else mrna.end
utr_end = max(utr_start + 1, mrna.end)
gene_utr_5primes.append(utr_start)
else:
utr_end = max(
mrna_utr_5primes) if mrna_utr_5primes else mrna.start
utr_start = min(mrna.start, utr_end - 1)
gene_utr_5primes.append(utr_end)
attr = mrna.attr.copy()
attr['Parent'] = attr['ID']
attr['ID'] = attr['ID'] + '-3UTR'
attr['color'] = '#008888'
utr = annotation.Annotation(
source='tt',
type='three_prime_utr',
seqid=mrna.seqid,
strand=mrna.strand,
start=utr_start,
end=utr_end,
attr=attr,
)
max_ext = _max_extension(utr, exon_index, mrna_end_index)
utr.attr["max_extension"] = str(max_ext)
#Only include if there is an annotated 3' UTR or end is not in the middle of some other isoform's exon
if utr_end - utr_start + max_ext > 1:
mrna_utrs.append(utr)
if gene.strand >= 0:
utr_start = max(
gene_utr_5primes) if gene_utr_5primes else gene.end
utr_end = max(utr_start + 1, gene.end)
else:
utr_end = min(
gene_utr_5primes) if gene_utr_5primes else gene.start
utr_start = min(gene.start, utr_end - 1)
attr = gene.attr.copy()
attr['Parent'] = attr['ID']
attr['ID'] = attr['ID'] + '-3UTR'
attr['color'] = '#008888'
utr = annotation.Annotation(
source='tt',
type='three_prime_utr',
seqid=gene.seqid,
strand=gene.strand,
start=utr_start,
end=utr_end,
attr=attr,
)
utr.attr["max_extension"] = str(
_max_extension(utr, exon_index, mrna_end_index))
gene_utrs.append(utr)
annotation.write_gff3(work / 'reference.gff', annotations + mrna_utrs)
annotation.write_gff3(work / 'utr.gff', gene_utrs)
work.update_param(tail_tools_reference_version=work.VERSION)
def run(self):
items = list(annotation.read_annotations(self.parent))
annotation.link_up_annotations(items)
for item in items:
assert len(item.parents) <= 1
genes = [ item for item in items if item.type == "gene" ]
downstrand_genes = [ _extend(_three_prime(item), self.extension) for item in genes ]
exons = [ item for item in items if item.type == "exon" ]
utrs = [ _extend(item, self.extension) for item in items if item.type == "three_prime_utr" ]
gene_index = span_index.index_annotations(genes)
downstrand_gene_index = span_index.index_annotations(downstrand_genes)
exon_index = span_index.index_annotations(exons)
utr_index = span_index.index_annotations(utrs)
peaks = [ ]
for peak in annotation.read_annotations(self.child):
if float(peak.attr.get("mean_tail","0.0")) < self.min_tail:
continue
peaks.append(peak)
for peak in peaks:
# Query is final base in genome before poly(A) starts
query = peak.three_prime().shifted(-1,0)
hit_to = "3'UTR"
hits = [ item.parents[0].parents[0] for item in
utr_index.get(query, True) ]
if not hits:
hit_to = "Exon"
hits = [ item.parents[0].parents[0] for item in
exon_index.get(query, True) ]
# For non-coding RNAs, which don't have a 3' UTR
if not hits:
hit_to = "Downstrand"
hits = downstrand_gene_index.get(query, True)
if not hits:
hit_to = "Intron"
hits = gene_index.get(query, True)
antisense_hits = gene_index.get(query.reversed(), True)
if not hits:
hit_to = "Antisense"
hits = antisense_hits
if hits:
peak.attr["Parent"] = join_descriptions([ item.get_id() for item in hits ], ",")
peak.attr["Relation"] = hit_to
peak.attr["Name"] = join_descriptions([ item.attr.get("Name","") for item in hits ])
peak.attr["Product"] = hit_to + " " + join_descriptions([ item.attr.get("Product","") for item in hits ])
peak.attr["Biotype"] = join_descriptions([ item.attr.get("Biotype","") for item in hits ])
if antisense_hits:
peak.attr["Antisense_parent"] = join_descriptions([ item.get_id() for item in antisense_hits ], ",")
peak.attr["Antisense_name"] = join_descriptions([ item.attr.get("Name","") for item in antisense_hits ])
peak.attr["Antisense_product"] = "Antisense " + join_descriptions([ item.attr.get("Product","") for item in antisense_hits ])
peak.attr["Antisense_biotype"] = join_descriptions([ item.attr.get("Biotype","") for item in antisense_hits])
annotation.write_gff3(self.prefix+"-parent.gff", genes) #Hmm
annotation.write_gff3(self.prefix+"-child.gff", peaks)
def run(self):
ref = env.load_ref(self.reference_dir)
analysis = env.load_analysis(self.analysis_dir)
samples = self.samples
if not samples:
samples = analysis.peak_counts['Count'].value_type().keys()
def total_count(peak_id):
return sum(analysis.peak_counts['Count'][peak_id][sample] for sample in samples)
#for each gene
#determine a cutoff region
#get the most highly expressed peak in that region
#output something useful
peaks_of = collections.defaultdict(list)
for peak in analysis.peaks.values():
if "Parent" in peak.attr:
peaks_of[peak.attr["Parent"]].append(peak)
called_peaks = [ ]
called_utrs = [ ]
called_genes = [ ]
n_good = 0
n_bad = 0
for utr in ref.utrs.values():
query = _extend(utr, self.extension)
candidates = [ ]
for peak in peaks_of[utr.attr["Parent"]]:
if peak.attr.get("Relation") != "3'UTR": continue
relpeak = peak.relative_to(query)
if relpeak.end <= 0: continue
candidates.append((-total_count(peak.get_id()), peak.get_id()))
candidates.sort()
if not candidates:
#print 'no peak for ', utr.attr['Parent']
n_bad += 1
continue
peak = analysis.peaks_asis[candidates[0][1]].copy()
peak.attr['Parent'] = utr.attr['Parent']
peak.attr['Name'] = utr.attr.get('Name','')
peak.attr['Product'] = utr.attr.get('Product','')
called_peaks.append(peak)
called_utr = utr.five_prime().span_with(peak.three_prime())
called_utr.attr['Peak'] = peak.get_id()
called_utrs.append(called_utr)
called_gene = ref.genes[utr.attr['Parent']].five_prime().span_with(peak.three_prime())
called_gene.attr['Peak'] = peak.get_id()
called_genes.append(called_gene)
n_good += 1
print n_good, 'UTR called'
print n_bad, 'no UTR called'
annotation.write_gff3(self.prefix + '-peaks.gff', called_peaks)
annotation.write_gff3(self.prefix + '-utrs.gff', called_utrs)
annotation.write_gff3(self.prefix + '-genes.gff', called_genes)
