def main(self):
for m in gff3.models(self.pre(sys.stdin)):
self.checkMiRnas(m)
self.checkPseudogene(m)
self.checkTranscriptNames(m)
for f in gff3.flattenModel(m):
print(str(f), end='')
def main():
log("Starting trimForAgr...")
#
soterm2id = loadSOTerms()
#
writeHeader([
('data-source', 'MGI'),
('date-produced', time.asctime()),
('assembly', os.environ["ENSEMBLbuild"]),
('annotationSource RefSeq', os.environ["NCBIver"]),
('annotationSource ENSEMBL', os.environ["ENSEMBLver"]),
])
#
for m in gff3.models(sys.stdin):
processModel(m, soterm2id)
def main(featureSources):
for fSource in featureSources:
for m in gff3.models(fSource):
stn = m.attributes.get("so_term_name",None)
count(m, [m.type+("[%s]"%stn if stn else "")], m)
#
pcounts("root", roots)
pcounts("mid", mids)
pcounts("leaf", leaves)
for k in exemplars:
es = list(exemplars[k])
es.sort()
if len(es) > 5:
es = es[::int(len(es)/5)]
paths[k] = str(paths[k])+ "\t" + "," .join(es)
pcounts("path", paths)
def loadPslFile(self):
# Each PSL line is parsed and turned into a gff3 feature hierarchy (match->match_part*)
# Here we iterate over the model roots (matches); the match_parts dangle below (f.children)
self.counts = {}
for m in gff3.models(psl.toGff(self.pslFile)):
#
seqid = m.qName.split(DOT)[0] # get the seqid w/o version number
mgiid = self.seqid2gene[seqid] # lookup the corresponding mgiid
mfeats = self.mgi2feats[
mgiid] # list containing the gene followed by its match features
if m.pctLength < MIN_PCT_LENGTH:
self.logRejects(
"REJECTING SEQUENCE (%s) for GENE (%s) - pctLength (%1.2f) less than minimum (%1.2f)"
% (seqid, mgiid, m.pctLength, MIN_PCT_LENGTH))
self.logRejects(str(m))
continue
mfeats.append(m)
self.counts[seqid] = self.counts.setdefault(seqid, 0) + 1
import sys
import gff3
for m in gff3.models(sys.stdin, flatten=True):
print(m)
import sys
import gff3
from OrderedSet import OrderedSet
EXCLUDE_SOURCES = OrderedSet(["NCBI"])
EXCLUDE_TYPES = OrderedSet([
"chromosome", "biological_region", "supercontig", "three_prime_UTR",
"five_prime_UTR"
])
filtFcn = lambda f: f.type not in EXCLUDE_TYPES and f.source not in EXCLUDE_SOURCES
feats = filter(filtFcn, gff3.iterate(sys.stdin))
for m in gff3.models(feats):
for f in gff3.flattenModel(m):
if f.attributes.get("ID", "").startswith("transcript:"):
f.Name = f.transcript_id
f.source = "ENSEMBL"
if len(f.parents) == 0:
f.attributes["curie"] = "ENSEMBL:" + f.ID.split(":")[1]
biotype = f.attributes.get("biotype", None)
if biotype and len(f.parents) == 0:
if biotype == "protein_coding":
biotype = "protein_coding_gene"
f.attributes["so_term_name"] = biotype
f.attributes.pop("biotype", None)
f.attributes.pop("version", None)
f.attributes.pop("description", None)
f.attributes.pop("logic_name", None)
f.attributes.pop("gene_id", None)
f.attributes.pop("transcript_support_level", None)
# canonicalize.py
#
# Turns tree models in canonical form DAG-shaped) models by merging of identical subfeatures.
#
import gff3
import sys
def mergeExon(ex, f):
ex.Parent.extend(f.Parent)
if f.source not in ex.source:
ex.source += ("," + f.source)
for feats in gff3.models(sys.stdin, flatten=True):
exons = {} # (start,end) -> index into ofeats
ofeats = [] # list of feats in current model. Root is 0th item.
# merge exons. exons merge if they have the same coordinates.
for f in feats:
if f.type == "exon":
k = (f.start, f.end)
if k in exons:
# merge f with a previously seen exon, ex
#
i = exons[k]
ex = ofeats[i]
mergeExon(ex, f)
#
# ...and move ex to end of the list
ofeats[i] = None