def main():
args = cli(sys.argv[0], sys.argv[1:])
if args.go is not None:
go = parse_rfam2go(args.go)
else:
go = {}
for line in args.infile:
if line.startswith("#"):
continue
record = GFFRecord.parse(line)
attrs = record.attributes
if args.best and attrs.custom["olp"] == "=":
continue
name = record.type
dbxrefs = ["Rfam:" + attrs.custom["mdlaccn"], "Rfam:" + name]
if "clan" in attrs.custom:
dbxrefs.append("RfamClan:" + attrs.custom["clan"])
ontology_terms = go.get(attrs.custom["mdlaccn"], [])
notes = [attrs.custom["desc"]]
target = Target(
attrs.custom["mdlaccn"],
int(attrs.custom["mdlfrom"]),
int(attrs.custom["mdlto"]),
)
custom = {
"evalue": attrs.custom["evalue"],
"model_type": attrs.custom["mdl"],
"gc": attrs.custom["gc"],
"bias": attrs.custom["bias"],
"bitscore": record.score,
}
if attrs.custom["trunc"] == "yes":
custom["truncated_match"] = "true"
if attrs.custom["olp"] == "=":
custom["overlap_with_better_score"] = "true"
record.source = args.source
record.type = args.type
record.score = float(custom["evalue"])
record.attributes = GFFAttributes(
name=name,
dbxref=dbxrefs,
target=target,
note=notes,
ontology_term=ontology_terms,
custom=custom,
)
print(record, file=args.outfile)
def decode_gff(infiles, outfile, map_, column):
inhandles = join_files(infiles, header=False)
if column == "id":
trans_function = replace_gff_id
elif column == "name":
trans_function = replace_gff_name
elif column == "seqid":
trans_function = replace_gff_seqid
else:
raise ValueError("This shouldn't ever happen")
record_chunk = list()
for i, line in enumerate(inhandles):
if line.startswith("#"):
record_chunk.append(line.strip())
continue
old_record = GFFRecord.parse(line)
new_records = trans_function(old_record, map_)
record_chunk.append(str(new_records))
if i % 10000 == 0:
outfile.write("\n".join(record_chunk))
record_chunk = list()
if len(record_chunk) > 0:
outfile.write("\n".join(record_chunk))
return
def encode_gff(
infiles,
outfile,
mapfile,
column,
id_conv,
):
inhandles = join_files(infiles, header=False)
seen = dict()
if column == "id":
trans_function = replace_gff_id
elif column == "name":
trans_function = replace_gff_name
elif column == "seqid":
trans_function = replace_gff_seqid
else:
raise ValueError("This shouldn't ever happen")
id_chunk = list()
record_chunk = list()
for i, line in enumerate(inhandles):
if line.startswith("#"):
record_chunk.append(line.strip())
continue
old_record = GFFRecord.parse(line)
new_record = trans_function(
old_record,
seen,
id_chunk,
lambda x: next(id_conv),
)
record_chunk.append(str(new_record))
if i % 10000 == 0:
if len(record_chunk) > 0:
outfile.write("\n".join(record_chunk))
outfile.write("\n")
record_chunk = list()
if len(id_chunk) > 0:
mapfile.write("".join(f"{n}\t{o}\n" for n, o in id_chunk))
id_chunk = list()
if len(record_chunk) > 0:
outfile.write("\n".join(record_chunk))
outfile.write("\n")
if len(id_chunk) > 0:
mapfile.write("".join(f"{n}\t{o}\n" for n, o in id_chunk))
return
def rnammer2gff(args: argparse.Namespace) -> None:
records: List[GFFRecord] = []
for line in args.infile:
if line.startswith("#"):
continue
sline = line.strip().split("\t")
rrna_type = sline[8]
new_type = TYPE_MAP[args.kingdom][rrna_type.lower()]
sline[1] = args.source
sline[2] = new_type
sline[8] = "."
rna_record = GFFRecord.parse("\t".join(sline))
gene_record = deepcopy(rna_record)
gene_record.type = "rRNA_gene"
gene_record.add_child(rna_record)
records.append(gene_record)
records.append(rna_record)
num = 0
for record in GFF(records).traverse_children(sort=True):
if record.attributes is None:
attr = GFFAttributes()
record.attributes = attr
else:
attr = record.attributes
if record.type == "rRNA_gene":
num += 1
attr.id = f"rRNA_gene{num}"
else:
attr.id = f"rRNA{num}"
attr.parent = [
p.attributes.id for p in record.parents
if (p.attributes is not None and p.attributes.id is not None)
]
print(record, file=args.outfile)
return
def main():
args = cli(sys.argv[0], sys.argv[1:])
rows = list()
for line in args.infile:
if line.startswith("#"):
continue
record = GFFRecord.parse(line)
record.attributes.id = record.attributes.name
record.attributes.name = None
record.attributes.custom = {}
rows.append(record)
gff = GFF(rows)
gff.infer_missing_parents()
counter = 1
for mrna in gff.select_type("mRNA"):
if len(mrna.children) < 2:
continue
region = deepcopy(mrna)
region.type = "repeat_region"
region.attributes.id = f"repeat_region{counter}"
region.attributes.ontology_term = ["SO:0000657"],
mrna.type = "helitron"
mrna.parents = [region]
mrna.attributes.parent = [region.attributes.id]
mrna.attributes.id = f"helitron{counter}"
mrna.attributes.ontology_term = ["SO:0000544", "SO:helitron"]
mrna.attributes.custom = {}
flank3 = [c for c in mrna.children
if c.attributes.id.endswith(".3")][0]
flank5 = [c for c in mrna.children
if c.attributes.id.endswith(".5.1")][0]
flank3.type = "three_prime_flanking_region"
flank3.attributes.ontology_term = [
"SO:0001417",
"SO:three_prime_flanking_region",
"SO:0000364",
"SO:transposable_element_flanking_region"
]
flank3.attributes.id = None
flank3.attributes.parent = [mrna.attributes.id]
flank5.type = "five_prime_flanking_region"
flank5.attributes.ontology_term = [
"SO:0001416",
"SO:five_prime_flanking_region",
"SO:0000364",
"SO:transposable_element_flanking_region"
]
flank5.attributes.id = None
flank5.attributes.parent = [mrna.attributes.id]
mrna.source = flank5.source
print(region, file=args.outfile)
print(mrna, file=args.outfile)
if mrna.strand == Strand.MINUS:
print(flank3, file=args.outfile)
print(flank5, file=args.outfile)
else:
print(flank5, file=args.outfile)
print(flank3, file=args.outfile)
counter += 1
return
def deal_with_block(block: List[str], gene_num: int) -> List[GFF3Record]:
parsed: Dict[str, List[GFFRecord[GTFAttributes]]] = dict()
for line in block:
rec = GFFRecord.parse(line, attr=GTFAttributes)
if rec.type in parsed:
parsed[rec.type].append(rec)
else:
parsed[rec.type] = [rec]
assert len(parsed["gene"]) == 1
assert len(parsed["similarity"]) == 1
gene_parsed = parsed["gene"][0]
similarity_parsed = parsed["similarity"][0]
custom: Dict[str, str] = dict()
if similarity_parsed.attributes is not None:
custom["query"] = similarity_parsed.attributes.custom["Query"]
if gene_parsed.attributes is not None:
custom["identity"] = gene_parsed.attributes.custom["identity"]
custom["similarity"] = gene_parsed.attributes.custom["similarity"]
gene = GFF3Record(
parsed["gene"][0].seqid,
"exonerate",
type="gene",
start=parsed["gene"][0].start,
end=parsed["gene"][0].end,
score=parsed["gene"][0].score,
strand=parsed["gene"][0].strand,
phase=parsed["gene"][0].phase,
attributes=GFF3Attributes(
id=f"gene{gene_num}",
custom=custom,
)
)
cdss = [
GFF3Record(
e.seqid,
"exonerate",
"CDS",
e.start,
e.end,
e.score,
e.strand,
e.phase,
attributes=GFF3Attributes(
id=f"CDS{gene_num}",
parent=[f"mRNA{gene_num}"],
custom=(e.attributes.custom
if e.attributes is not None
else None)
)
)
for e
in parsed["exon"]
]
for c in cdss:
if gene.attributes is not None:
# This is safe because we added attributes.
assert c.attributes is not None
c.attributes.custom["query"] = gene.attributes.custom["query"]
mrna = GFF3Record.infer_from_children(
cdss,
id=f"mRNA{gene_num}",
seqid=gene.seqid,
source="exonerate",
type="mRNA",
strand=gene.strand,
score=gene.score,
)
mrna.add_parent(gene)
if gene.attributes is not None:
# This is safe because infer_from_children adds an ID to attributes.
assert mrna.attributes is not None
if gene.attributes.id is not None:
mrna.attributes.parent = [gene.attributes.id]
mrna.attributes.custom["query"] = gene.attributes.custom["query"]
out = [gene, mrna]
out.extend(cdss)
return out