def main():
parser = argparse.ArgumentParser(
description=
'Updates exon Parent attributes to point at the correct RNA feature')
## output file to be written
parser.add_argument('-i',
'--input',
type=str,
required=True,
help='Path to the input GFF3 file')
parser.add_argument('-o',
'--output',
type=str,
required=True,
help='Output GFF3 file to write')
args = parser.parse_args()
infile = open(args.input)
ofh = open(args.output, 'wt')
last_rna_id = None
for line in infile:
if line.startswith('#'):
ofh.write(line)
continue
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
ofh.write("{0}\n".format(line))
continue
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
if cols[2].endswith('RNA'):
last_rna_id = id
ofh.write("{0}\n".format(line))
elif cols[2] == 'exon':
if parent != last_rna_id:
print(
"INFO: correcting unexpected parentage for feature ({0}) type {2}. Expected ({1})"
.format(id, last_rna_id, cols[2]))
cols[8] = gff.set_column_9_value(cols[8], 'Parent',
last_rna_id)
ofh.write("{0}\n".format("\t".join(cols)))
else:
ofh.write("{0}\n".format(line))
else:
ofh.write("{0}\n".format(line))
def main():
flawed_gff_file = 'canonical.flawed.gff3'
ilri_gff = 'Theileria-all-Theileria1_ourids.gff'
source = 'GenBank'
out_gff = 'canonical.corrected.gff3'
fout = open(out_gff, mode='wt', encoding='utf-8')
fout.write("##gff-version 3\n")
(assemblies, features) = gff.get_gff3_features(flawed_gff_file)
print("INFO: loaded {0} assemblies and {1} features".format(len(assemblies), len(features)))
polypeptides = dict()
for line in open(ilri_gff):
cols = line.split("\t")
if len(cols) != 9 or cols[2] != 'polypeptide':
continue
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
polypeptides[parent] = things.Polypeptide(id=id, parent=parent)
polypeptides[parent].locate_on(target=assemblies[cols[0]], fmin=int(cols[3]) - 1, fmax=int(cols[4]), strand=cols[6])
print("DEBUG: loaded {0} polypeptides from ILRI file".format(len(polypeptides)) )
for assembly_id in assemblies:
for gene in assemblies[assembly_id].genes():
for mRNA in gene.mRNAs():
if mRNA.id not in polypeptides:
print("DEBUG: {0} not found as a parent to any polypeptide".format(mRNA.id))
else:
polypeptide = polypeptides[mRNA.id]
# pull this outside of the iteration since iterating might delete some
CDSs = mRNA.CDSs()
for CDS in CDSs:
keep = True
if CDS < polypeptide:
mRNA.delete_CDS(CDS)
elif CDS <= polypeptide:
CDS.location().fmin = polypeptide.location().fmin
if CDS > polypeptide:
mRNA.delete_CDS(CDS)
elif CDS >= polypeptide:
CDS.location().fmax = polypeptide.location().fmax
#print("WARN: found a CDS {0}:{1}-{2} outside the range of the polypeptide {3}:{4}-{5}".format( \
# CDS.id, CDS.locations[0].fmin, CDS.locations[0].fmax, \
# polypeptide.id, polypeptide.locations[0].fmin, polypeptide.locations[0].fmax))
gene.print_as(fh=fout, source=source, format='gff3')
def main():
parser = argparse.ArgumentParser( description='Generates new identifiers in GFF3 files following the IGS identifier convention.')
## output file to be written
parser.add_argument('-i', '--input_file', type=str, required=True, help='TA file of source molecules' )
parser.add_argument('-o', '--output_file', type=str, required=False, help='Optional output file path (else STDOUT)' )
parser.add_argument('-p', '--prefix', type=str, required=True, help='The prefix portion of IDs to be generated')
parser.add_argument('-m', '--mode', type=str, required=False, default='sequential', help='ID modes (see embedded documentation): sequential, uuid, hex8, hex12')
args = parser.parse_args()
check_arguments(args)
id_map = dict()
## output will either be a file or STDOUT
fout = sys.stdout
if args.output_file is not None:
fout = open(args.output_file, 'wt')
for line in open(args.input_file):
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
fout.write(line + "\n")
continue
# grab the ID column if any
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
new_id = None
new_parent = None
type = cols[2]
if id is not None:
if id in id_map:
new_id = id_map[id]
else:
new_id = get_new_id(args.prefix, type, args.mode)
id_map[id] = new_id
cols[8] = cols[8].replace("ID={0}".format(id), "ID={0}".format(new_id))
if parent is not None:
if parent in id_map:
new_parent = id_map[parent]
else:
raise Exception("ERROR: parent ({0}) referenced before it was used as an ID".format(parent))
cols[8] = cols[8].replace("Parent={0}".format(parent), "Parent={0}".format(new_parent))
#print("DEBUG: old_id:{0} - old_parent:{1}, new_id:{2} - new_parent:{3}".format(id, parent, new_id, new_parent))
fout.write("\t".join(cols) + "\n")
def main():
parser = argparse.ArgumentParser('Filter the genes of a GFF3 file by mRNA child IDs')
## output file to be written
parser.add_argument('-i', '--input_gff', type=str, required=True, help='GFF file of source annotation' )
parser.add_argument('-o', '--output_file', type=str, required=False, help='Optional output file path (else STDOUT)' )
args = parser.parse_args()
## output will either be a file or STDOUT
fout = sys.stdout
if args.output_file is not None:
fout = open(args.output_file, 'wt')
current_mRNA_id = None
current_mol_id = None
current_fragments = list()
current_direction = None
for line in open(args.input_gff):
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
continue
# grab the ID and Parent columns if any
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
mol_id = cols[0]
type = cols[2]
if type == 'mRNA':
if current_mRNA_id is not None and id != current_mRNA_id:
# purge the existing one first
write_transcript(fout, current_mol_id, current_fragments, current_direction)
current_fragments = list()
current_mRNA_id = id
current_mol_id = cols[0]
current_direction = cols[6]
elif type == 'exon':
if cols[6] == '+':
current_fragments.append({'start':cols[3], 'end':cols[4]})
else:
current_fragments.append({'start':cols[4], 'end':cols[3]})
write_transcript(fout, current_mol_id, current_fragments, current_direction)
def main():
parser = argparse.ArgumentParser(
description='Adds gene features for RNAs which lack them')
## output file to be written
parser.add_argument('-i',
'--input',
type=str,
required=True,
help='Path to the input GFF3 file')
parser.add_argument('-o',
'--output',
type=str,
required=True,
help='Output GFF3 file to write')
args = parser.parse_args()
infile = open(args.input)
ofh = open(args.output, 'wt')
for line in infile:
if line.startswith('#'):
ofh.write(line)
continue
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
ofh.write("{0}\n".format(line))
continue
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
if cols[2].endswith('RNA') and parent is None:
gene_cols = list(cols)
gene_cols[2] = 'gene'
gene_cols[8] = gff.set_column_9_value(gene_cols[8], 'ID',
"{0}.gene".format(id))
ofh.write("{0}\n".format("\t".join(gene_cols)))
cols[8] = gff.set_column_9_value(cols[8], 'Parent',
"{0}.gene".format(id))
ofh.write("{0}\n".format("\t".join(cols)))
else:
ofh.write("{0}\n".format(line))
def main():
parser = argparse.ArgumentParser( description='Put a description of your script here')
## output file to be written
parser.add_argument('-i', '--input_file', type=str, required=True, help='Path to an input file to be read' )
parser.add_argument('-o', '--output_gff', type=str, required=True, help='Path to an output GFF file to be created with new IDs' )
parser.add_argument('-p', '--id_prefix', type=str, required=True, help='Will be used as the base for all IDs generated' )
parser.add_argument('-m', '--output_map', type=str, required=False, help='This will create a tab-delimited mapping of old and new IDs' )
args = parser.parse_args()
ofh = open(args.output_gff, 'w')
if args.output_map is None:
map_ofh = None
else:
map_ofh = open(args.output_map, 'w')
idmap = dict()
for line in open(args.input_file):
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
ofh.write(line + "\n")
continue
feat_id = gff.column_9_value(cols[8], 'ID')
parent_id = gff.column_9_value(cols[8], 'Parent')
if feat_id in idmap:
new_feat_id = idmap[feat_id]
else:
new_feat_id = get_new_id(args.id_prefix, cols[2], feat_id, map_ofh)
idmap[feat_id] = new_feat_id
if parent_id is None:
cols[8] = "ID={0}".format(new_feat_id)
else:
if parent_id in idmap:
new_parent_id = idmap[parent_id]
else:
new_parent_id = get_new_id(args.id_prefix, cols[2], parent_id, map_ofh)
idmap[parent_id] = new_parent_id
cols[8] = "ID={0};Parent={1}".format(new_feat_id, new_parent_id)
ofh.write( "\t".join(cols) + "\n" )
def append_organism_names_to_gff(file_path, poly_orgs):
# we have to write to a temp file and copy over
fout = open("{0}.orgtmp".format(file_path), 'wt')
orgs_found = 0
last_RNA_id = None
for line in open(file_path):
line = line.rstrip()
cols = line.split("\t")
if len(cols) == 9 and cols[2].endswith('RNA'):
last_RNA_id = gff.column_9_value(cols[8], 'ID')
if len(cols) == 9 and cols[2] == 'polypeptide':
if last_RNA_id in poly_orgs:
cols[8] += ";top_organism_from_blast={0}".format(
poly_orgs[last_RNA_id], gff.escape(poly_orgs[last_RNA_id]))
orgs_found += 1
fout.write("{0}\n".format("\t".join(cols)))
else:
fout.write("{0}\n".format(line))
if orgs_found == 0:
print(
"WARNING: The --export_organism_names option was passed, but parsing failed to find any organism names at all. This might be an error."
)
## now move the temp file over the original copy
fout.close()
os.rename("{0}.orgtmp".format(file_path), file_path)
def append_organism_names_to_gff(file_path, poly_orgs):
# we have to write to a temp file and copy over
fout = open("{0}.orgtmp".format(file_path), 'wt')
orgs_found = 0
last_RNA_id = None
for line in open(file_path):
line = line.rstrip()
cols = line.split("\t")
if len(cols) == 9 and cols[2].endswith('RNA'):
last_RNA_id = gff.column_9_value(cols[8], 'ID')
if len(cols) == 9 and cols[2] == 'polypeptide':
if last_RNA_id in poly_orgs:
cols[8] += ";top_organism_from_blast={0}".format(poly_orgs[last_RNA_id], gff.escape(poly_orgs[last_RNA_id]))
orgs_found += 1
fout.write("{0}\n".format("\t".join(cols)) )
else:
fout.write("{0}\n".format(line))
if orgs_found == 0:
print("WARNING: The --export_organism_names option was passed, but parsing failed to find any organism names at all. This might be an error.")
## now move the temp file over the original copy
fout.close()
os.rename("{0}.orgtmp".format(file_path), file_path)
def main():
parser = argparse.ArgumentParser( description='Reverses CDS coodinates where stop < start')
## output file to be written
parser.add_argument('-i', '--input', type=str, required=True, help='Path to the input GFF3 file' )
parser.add_argument('-o', '--output', type=str, required=True, help='Output GFF3 file to write' )
args = parser.parse_args()
infile = open(args.input)
ofh = open(args.output, 'wt')
for line in infile:
if line.startswith('#'):
ofh.write(line)
continue
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
ofh.write("{0}\n".format(line) )
continue
if cols[2] == 'CDS' and int(cols[4]) < int(cols[3]):
temp = cols[3]
cols[3] = cols[4]
cols[4] = temp
id = gff.column_9_value(cols[8], 'ID')
print("CDS reversed: {0}".format(id))
ofh.write("{0}\n".format("\t".join(cols)) )
else:
ofh.write("{0}\n".format(line) )
def main():
parser = argparse.ArgumentParser( description='Updates exon Parent attributes to point at the correct RNA feature')
## output file to be written
parser.add_argument('-i', '--input', type=str, required=True, help='Path to the input GFF3 file' )
parser.add_argument('-o', '--output', type=str, required=True, help='Output GFF3 file to write' )
args = parser.parse_args()
infile = open(args.input)
ofh = open(args.output, 'wt')
last_rna_id = None
for line in infile:
if line.startswith('#'):
ofh.write(line)
continue
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
ofh.write("{0}\n".format(line) )
continue
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
if cols[2].endswith('RNA'):
last_rna_id = id
ofh.write("{0}\n".format(line) )
elif cols[2] == 'exon':
if parent != last_rna_id:
print("INFO: correcting unexpected parentage for feature ({0}) type {2}. Expected ({1})".format(id, last_rna_id, cols[2]) )
cols[8] = gff.set_column_9_value(cols[8], 'Parent', last_rna_id)
ofh.write("{0}\n".format("\t".join(cols)) )
else:
ofh.write("{0}\n".format(line) )
else:
ofh.write("{0}\n".format(line) )
def main():
parser = argparse.ArgumentParser( description='Adds gene features for RNAs which lack them')
## output file to be written
parser.add_argument('-i', '--input', type=str, required=True, help='Path to the input GFF3 file' )
parser.add_argument('-o', '--output', type=str, required=True, help='Output GFF3 file to write' )
args = parser.parse_args()
infile = open(args.input)
ofh = open(args.output, 'wt')
for line in infile:
if line.startswith('#'):
ofh.write(line)
continue
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
ofh.write("{0}\n".format(line) )
continue
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
if cols[2].endswith('RNA') and parent is None:
gene_cols = list(cols)
gene_cols[2] = 'gene'
gene_cols[8] = gff.set_column_9_value(gene_cols[8], 'ID', "{0}.gene".format(id))
ofh.write("{0}\n".format("\t".join(gene_cols)) )
cols[8] = gff.set_column_9_value(cols[8], 'Parent', "{0}.gene".format(id))
ofh.write("{0}\n".format("\t".join(cols)) )
else:
ofh.write("{0}\n".format(line) )
def main():
parser = argparse.ArgumentParser( description='Removes duplicate features in a GFF3 file')
## output file to be written
parser.add_argument('-i', '--input', type=str, required=True, help='Path to the input GFF3 file' )
parser.add_argument('-o', '--output', type=str, required=True, help='Output GFF3 file to write' )
args = parser.parse_args()
# just reduce the keys to a string:
# "molecule__parent__type__start__stop"
found = list()
infile = open(args.input)
outfile = open(args.output, 'wt')
for line in infile:
if line.startswith('#'):
outfile.write(line)
continue
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
continue
parent = gff.column_9_value(cols[8], 'Parent')
type = cols[2]
mol_id = cols[0]
if parent is None:
outfile.write("{0}\n".format(line))
continue
id_string = "{0}__{1}__{2}__{3}__{4}".format(mol_id, parent, type, cols[3], cols[4])
if id_string in found:
print("INFO: duplicate feature to be removed:\n{0}\n".format(line) )
continue
else:
found.append(id_string)
outfile.write("{0}\n".format(line) )
def main():
parser = argparse.ArgumentParser(
description='Reverses CDS coodinates where stop < start')
## output file to be written
parser.add_argument('-i',
'--input',
type=str,
required=True,
help='Path to the input GFF3 file')
parser.add_argument('-o',
'--output',
type=str,
required=True,
help='Output GFF3 file to write')
args = parser.parse_args()
infile = open(args.input)
ofh = open(args.output, 'wt')
for line in infile:
if line.startswith('#'):
ofh.write(line)
continue
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
ofh.write("{0}\n".format(line))
continue
if cols[2] == 'CDS' and int(cols[4]) < int(cols[3]):
temp = cols[3]
cols[3] = cols[4]
cols[4] = temp
id = gff.column_9_value(cols[8], 'ID')
print("CDS reversed: {0}".format(id))
ofh.write("{0}\n".format("\t".join(cols)))
else:
ofh.write("{0}\n".format(line))
def main():
parser = argparse.ArgumentParser( description='Convert native (GTF) or GFF output from Augustus into GFF3 format')
## output file to be written
parser.add_argument('-i', '--input', type=str, required=True, help='Path to a GFF file created by Augustus' )
parser.add_argument('-o', '--output', type=str, required=True, help='Path to an output file to be created' )
args = parser.parse_args()
assemblies = dict()
current_assembly = None
gene = None
mRNAs = dict()
in_sequence = False
current_sequence = None
current_gene_comment_lines = list()
## Used for tracking the exon count for each gene (for ID purposes)
exon_count_by_mRNA = dict()
fout = open(args.output, mode='wt', encoding='utf-8')
fout.write("##gff-version 3\n")
for line in open(args.input):
if line.startswith("#"):
current_gene_comment_lines.append(line)
if line.startswith("# end gene "):
## purge the comments, then write the gene
fout.write( "".join(current_gene_comment_lines) )
gene.print_as(fh=fout, source='AUGUSTUS', format='gff3')
gene = None
mRNAs = dict()
in_sequence = False
current_sequence = None
current_gene_comment_lines = list()
elif line.startswith("# protein sequence = ["):
pass
elif in_sequence is True:
# build 'current_sequence'
pass
else:
cols = line.split("\t")
if len(cols) != 9:
continue
mol_id = cols[0]
feat_type = cols[2]
if feat_type not in ['gene', 'transcript', 'CDS']:
continue
## The output format is GTF by default and (mostly) GFF if the --gff option is used.
# If GTF is detected, let's start by transforming the 9th column into GFF so the
# libraries can use it
# g1 -> ID=g1
# g1.t1 -> ID=g1.t1;Parent=g1
# transcript_id "g1.t1"; gene_id "g1"; -> ID=g1.t1.cds;Parent=g1.t1
m_gene = re.match('(g\d+)', cols[8])
m_transcript = re.match('((g\d+).t\d+)', cols[8])
m_CDS = re.match('transcript_id "(g\d+.t\d+)"; gene_id "g\d+";', cols[8])
# the input can be in GTF or GFF. We need to reformat the 9th column for the GTF entries
if not cols[8].startswith('ID') and not cols[8].startswith('Parent'):
if feat_type == 'gene':
if m_gene:
cols[8] = "ID={0}".format(m_gene.group(1))
else:
raise Exception("ERROR: GTF detected but gene row has bad 9th column format: {0}".format(cols[8]))
elif feat_type == 'transcript':
if m_transcript:
cols[8] = "ID={0};Parent={1}".format(m_transcript.group(1), m_transcript.group(2))
else:
raise Exception("ERROR: GTF detected but transcript row has bad 9th column format: {0}".format(cols[8]))
elif feat_type == 'CDS':
if m_CDS:
cols[8] = "ID={0}.cds;Parent={0}".format(m_CDS.group(1))
else:
raise Exception("ERROR: GTF detected but CDS row has bad 9th column format: {0}".format(cols[8]))
feat_id = gff.column_9_value(cols[8], 'ID')
## initialize this assembly if we haven't seen it yet
if mol_id not in assemblies:
assemblies[mol_id] = things.Assembly(id=mol_id)
current_assembly = assemblies[mol_id]
if feat_type == "gene":
gene = things.Gene(id=feat_id)
gene.locate_on( target=current_assembly, fmin=int(cols[3]) - 1, fmax=int(cols[4]), strand=cols[6] )
elif feat_type == "transcript":
mRNA = things.mRNA(id=feat_id, parent=gene)
mRNA.locate_on( target=current_assembly, fmin=int(cols[3]) - 1, fmax=int(cols[4]), strand=cols[6] )
gene.add_mRNA(mRNA)
mRNAs[mRNA.id] = mRNA
if feat_id in exon_count_by_mRNA:
raise Exception( "ERROR: two different mRNAs found with same ID: {0}".format(feat_id) )
else:
exon_count_by_mRNA[feat_id] = 0
elif feat_type == "CDS":
parent_id = gff.column_9_value(cols[8], 'Parent')
## sanity check that we've seen this parent
if parent_id not in mRNAs:
raise Exception("ERROR: Found CDS column with parent ({0}) mRNA not yet in the file".format(parent_id))
CDS = things.CDS(id=feat_id, parent=mRNAs[parent_id])
CDS.locate_on( target=current_assembly, fmin=int(cols[3]) - 1, fmax=int(cols[4]), strand=cols[6], phase=int(cols[7]) )
mRNA.add_CDS(CDS)
## exons weren't explicitly defined in the input file, so we need to derive new IDs for them
exon_count_by_mRNA[parent_id] += 1
exon_id = "{0}.exon{1}".format(parent_id, exon_count_by_mRNA[parent_id])
exon = things.Exon(id=exon_id, parent=mRNAs[parent_id])
exon.locate_on( target=current_assembly, fmin=int(cols[3]) - 1, fmax=int(cols[4]), strand=cols[6] )
mRNA.add_exon(exon)
def main():
parser = argparse.ArgumentParser(description="Adds locus tag identifiers to GFF3 features")
## output file to be written
parser.add_argument("-i", "--input_file", type=str, required=True, help="TA file of source molecules")
parser.add_argument("-o", "--output_file", type=str, required=False, help="Optional output file path (else STDOUT)")
parser.add_argument("-p", "--prefix", type=str, required=True, help="The prefix portion of IDs to be generated")
parser.add_argument(
"-a",
"--padding",
type=int,
required=True,
help="Specify the minimum with to reserve for the numeric portion of the IDs. Smaller numbers will be zero-padded.",
)
parser.add_argument(
"-n", "--interval", type=int, required=False, default=1, help="Interval between generated identifiers"
)
parser.add_argument(
"-s",
"--starting_id",
type=int,
required=False,
default=0,
help="Initial numeric portion of IDs to be generated (do not zero-pad)",
)
parser.add_argument(
"-d",
"--id_file",
type=str,
required=False,
help="Pass a 2-column file of IDs to retain (in case you have mapped genes, for example)",
)
parser.add_argument(
"-m",
"--molecule_map",
type=str,
required=False,
help="Pass a 2-column file of molecule->token identifiers (see documentation)",
)
parser.add_argument(
"-c", "--custom", type=str, required=False, help="For custom parsing steps. Most should ignore this."
)
args = parser.parse_args()
check_arguments(args)
# used to store locus_tags associated with each gene (so children can inherit)
gene_loci = dict()
next_id = args.starting_id
last_molecule = None
id_mapping = parse_mapping_file(args.id_file)
mol_mapping = parse_mapping_file(args.molecule_map)
loci_assigned = list()
## if using Joana's custom options, check assumptions
if args.custom == "joana":
if args.molecule_map is None or args.id_file is None:
raise Exception("ERROR: Expected --molecule_map and --id_file options when using --custom=joana")
else:
## need to process the ID map to reformat IDs
for id in id_mapping:
# TP05_0002 -> TpMuguga_05g00002
m = re.match("TP(\d\d)_(\d+)", id_mapping[id])
if m:
id_mapping[id] = "{0}_{1}g0{2}".format(args.prefix, m.group(1), m.group(2))
elif args.custom == "bmicroti":
microti_map = {"I": "01", "II": "02", "III": "03", "IV": "04"}
if args.molecule_map is None or args.id_file is None:
raise Exception("ERROR: Expected --molecule_map and --id_file options when using --custom=bmicroti")
else:
for id in id_mapping:
m = re.match("BBM_(\D+)(\d+)", id_mapping[id])
if m:
print("Changing id from {0} to ".format(id))
id_mapping[id] = "{0}_{1}g{2}".format(args.prefix, microti_map[m.group(1)], m.group(2))
print(id_mapping[id])
else:
raise Exception("ERROR: id ({0}) didn't match expected convention.".format(id_mapping[id]))
## output will either be a file or STDOUT
fout = sys.stdout
if args.output_file is not None:
fout = open(args.output_file, "wt")
last_number_portion_assigned = 0
for line in open(args.input_file):
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
fout.write(line + "\n")
continue
if last_molecule is None or (
args.molecule_map is not None and mol_mapping[cols[0]] != mol_mapping[last_molecule]
):
print("Found molecule {0}, resetting id counter from {1}".format(cols[0], next_id))
next_id = args.starting_id
last_molecule = cols[0]
# grab the ID column if any
id = gff.column_9_value(cols[8], "ID")
parent = gff.column_9_value(cols[8], "Parent")
type = cols[2]
# issue
# 66F4EEF2E3C863C251F831817FF71233
# 7F1917E4D81A959078C9A38E15488BC0
# E22888670919A4A888572155F40F2654
# B9D9CF1F7A8E5A2E1124F0A6C68840DC -> BBM_I00232
# gene before is: 6DE6BCCE69CCDC39994A0940B2ED524A - novel
# errors on: BmicrotiR1_01g00233 -> BBM_I00233
# 5800A4110A62E4EAE57AFAD1F8D65CB3 BBM_I00233
if type == "gene":
while True:
if id in id_mapping:
locus_id = id_mapping[id]
else:
if args.molecule_map is None:
locus_id = "{0}_{1}".format(args.prefix, str(next_id).zfill(args.padding))
else:
if cols[0] in mol_mapping:
if args.custom == "bmicroti":
locus_id = "{0}_{2}g{1}".format(
args.prefix,
str(int(last_number_portion_assigned) + 1).zfill(args.padding),
mol_mapping[cols[0]],
)
else:
locus_id = "{0}_{2}g{1}".format(
args.prefix, str(next_id).zfill(args.padding), mol_mapping[cols[0]]
)
else:
raise Exception("ERROR: --molecule_map passed but {0} wasn't found in it.".format(cols[0]))
next_id += args.interval
cols[8] = gff.set_column_9_value(cols[8], "locus_tag", locus_id)
## make sure this wasn't generated already (possibly conflict between --id_file and an
# auto-generated ID?
if locus_id not in loci_assigned:
break
else:
print("DEBUG: Duplicate ID assigned ({0}), trying again.".format(locus_id))
loci_assigned.append(locus_id)
gene_loci[id] = locus_id
m = re.search(r"(\d+)$", locus_id)
if m:
last_number_portion_assigned = m.group(1)
elif type.endswith("RNA"):
if parent in gene_loci:
cols[8] = gff.set_column_9_value(cols[8], "locus_tag", gene_loci[parent])
else:
raise Exception("ERROR: found RNA {0} whose parent {1} wasn't found yet".format(id, parent))
fout.write("\t".join(cols) + "\n")
def main():
parser = argparse.ArgumentParser( description='A GTF -> GFF3 conversion script for Cufflinks output')
## output file to be written
parser.add_argument('-i', '--input_file', type=str, required=True, help='Path to an input GTF file' )
parser.add_argument('-o', '--output_file', type=str, required=False, help='Path to an output GFF file to be created' )
parser.add_argument('-e', '--export_mode', type=str, required=False, default='model', help='Export mode for results (model or cDNA_match)' )
args = parser.parse_args()
if args.export_mode not in ['model', 'cDNA_match']:
raise Exception("ERROR: the only valid values for --export_mode are 'model' or 'cDNA_match'")
## output will either be a file or STDOUT
ofh = sys.stdout
if args.output_file is not None:
ofh = open(args.output_file, 'wt')
ofh.write("##gff-version 3\n")
assemblies = dict()
current_assembly = None
current_gene = None
current_RNA = None
current_match = None
rna_count_by_gene = defaultdict(int)
exon_count_by_RNA = defaultdict(int)
# each gb_record is a SeqRecord object
for line in open(args.input_file, "r"):
cols = line.split("\t")
if len(cols) != 9:
print("SKIPPING: {0}".format(line))
continue
mol_id = cols[0]
if mol_id not in assemblies:
assemblies[mol_id] = things.Assembly(id=mol_id)
current_assembly = assemblies[mol_id]
ftype = cols[2]
fmin = int(cols[3]) - 1
fmax = int(cols[4])
strand = cols[6]
col9 = cols[8]
# this makes it look like GFF column 9 so I can use biocodeutils.column_9_value(str, key)
col9 = col9.replace(' "', '="')
gene_id = gff.column_9_value(col9, 'gene_id').replace('"', '')
transcript_id = gff.column_9_value(col9, 'transcript_id').replace('"', '')
if ftype == 'transcript':
if args.export_mode == 'model':
if current_gene is not None and current_gene.id != gene_id:
gene.print_as(fh=ofh, source='Cufflinks', format='gff3')
if current_gene is None or current_gene.id != gene_id:
gene = things.Gene(id=gene_id)
gene.locate_on( target=current_assembly, fmin=fmin, fmax=fmax, strand=strand )
current_gene = gene
mRNA = things.mRNA(id=transcript_id, parent=current_gene)
mRNA.locate_on( target=current_assembly, fmin=fmin, fmax=fmax, strand=strand )
gene.add_mRNA(mRNA)
current_RNA = mRNA
exon_count_by_RNA[transcript_id] = 0
current_CDS_phase = 0
elif args.export_mode == 'cDNA_match':
if current_match is not None and current_match.id != transcript_id:
match.print_as( fh=ofh, source='Cufflinks', format='gff3' )
match = things.Match(id=transcript_id, subclass='cDNA_match', length=fmax - fmin)
match.locate_on( target=current_assembly, fmin=fmin, fmax=fmax, strand=strand )
current_match = match
elif ftype == 'exon':
exon_number = gff.column_9_value(col9, 'exon_number').replace('"', '')
if args.export_mode == 'model':
exon_count_by_RNA[transcript_id] += 1
cds_id = "{0}.CDS.{1}".format( current_RNA.id, exon_count_by_RNA[current_RNA.id] )
CDS = things.CDS(id=cds_id, parent=current_RNA)
CDS.locate_on( target=current_assembly, fmin=fmin, fmax=fmax, strand=strand, phase=current_CDS_phase )
current_RNA.add_CDS(CDS)
# calculate the starting phase for the next CDS feature (in case there is one)
current_CDS_phase = 3 - (((fmax - fmin) - current_CDS_phase) % 3)
if current_CDS_phase == 3:
current_CDS_phase = 0
exon_id = "{0}.exon.{1}".format( current_RNA.id, exon_count_by_RNA[current_RNA.id] )
exon = things.Exon(id=exon_id, parent=current_RNA)
exon.locate_on( target=current_assembly, fmin=fmin, fmax=fmax, strand=strand )
current_RNA.add_exon(exon)
elif args.export_mode == 'cDNA_match':
mp_id = "{0}.match_part.{1}".format(transcript_id, exon_number)
mp = things.MatchPart(id=mp_id, parent=current_match, length=fmax - fmin)
mp.locate_on( target=current_assembly, fmin=fmin, fmax=fmax, strand=strand )
current_match.add_part(mp)
# don't forget to do the last gene, if there were any
if args.export_mode == 'model':
if current_gene is not None:
gene.print_as(fh=ofh, source='GenBank', format='gff3')
elif args.export_mode == 'cDNA_match':
if current_match is not None:
match.print_as( fh=ofh, source='Cufflinks', format='gff3' )
def main():
parser = argparse.ArgumentParser(
description='Removes orphaned features in a GFF3 file')
## output file to be written
parser.add_argument('-i',
'--input',
type=str,
required=True,
help='Path to the input GFF3 file')
parser.add_argument('-o',
'--output',
type=str,
required=True,
help='Output GFF3 file to write')
#parser.add_argument('-t', '--type', type=str, required=False, help='Type of features to remove' )
args = parser.parse_args()
# going to try saving memory by tracking line numbers instead of storing all of it
# true means keep the line, false means to omit it
# doing tracking this way since it's technically legal for a feature to have no identifier at all.
lines = list()
parents = dict()
current_line_num = -1
infile = open(args.input)
for line in infile:
current_line_num += 1
if line.startswith('#'):
lines.append(True)
continue
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
lines.append(True)
continue
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
if parent is None:
# this might be overwritten later
lines.append(False)
if id is not None:
if parent not in parents:
parents[parent] = False
else:
lines.append(True)
parents[parent] = True
infile.seek(0)
current_line_num = -1
outfh = open(args.output, 'wt')
for line in infile:
current_line_num += 1
if lines[current_line_num] == True:
outfh.write(line)
else:
line = line.rstrip()
cols = line.split("\t")
if len(cols) == 9:
id = gff.column_9_value(cols[8], 'ID')
if id is not None and id in parents and parents[id] == True:
outfh.write("{0}\n".format(line))
else:
print("WARN: removing this line: {0}".format(line))
def main():
parser = argparse.ArgumentParser(
description='A GTF -> GFF3 conversion script for Cufflinks output')
## output file to be written
parser.add_argument('-i',
'--input_file',
type=str,
required=True,
help='Path to an input GTF file')
parser.add_argument('-o',
'--output_file',
type=str,
required=False,
help='Path to an output GFF file to be created')
parser.add_argument('-e',
'--export_mode',
type=str,
required=False,
default='model',
help='Export mode for results (model or cDNA_match)')
args = parser.parse_args()
if args.export_mode not in ['model', 'cDNA_match']:
raise Exception(
"ERROR: the only valid values for --export_mode are 'model' or 'cDNA_match'"
)
## output will either be a file or STDOUT
ofh = sys.stdout
if args.output_file is not None:
ofh = open(args.output_file, 'wt')
ofh.write("##gff-version 3\n")
assemblies = dict()
current_assembly = None
current_gene = None
current_RNA = None
current_match = None
rna_count_by_gene = defaultdict(int)
exon_count_by_RNA = defaultdict(int)
# each gb_record is a SeqRecord object
for line in open(args.input_file, "r"):
cols = line.split("\t")
if len(cols) != 9:
print("SKIPPING: {0}".format(line))
continue
mol_id = cols[0]
if mol_id not in assemblies:
assemblies[mol_id] = things.Assembly(id=mol_id)
current_assembly = assemblies[mol_id]
ftype = cols[2]
fmin = int(cols[3]) - 1
fmax = int(cols[4])
strand = cols[6]
col9 = cols[8]
# this makes it look like GFF column 9 so I can use biocodeutils.column_9_value(str, key)
col9 = col9.replace(' "', '="')
gene_id = gff.column_9_value(col9, 'gene_id').replace('"', '')
transcript_id = gff.column_9_value(col9,
'transcript_id').replace('"', '')
if ftype == 'transcript':
if args.export_mode == 'model':
if current_gene is not None and current_gene.id != gene_id:
gene.print_as(fh=ofh, source='Cufflinks', format='gff3')
if current_gene is None or current_gene.id != gene_id:
gene = things.Gene(id=gene_id)
gene.locate_on(target=current_assembly,
fmin=fmin,
fmax=fmax,
strand=strand)
current_gene = gene
mRNA = things.mRNA(id=transcript_id, parent=current_gene)
mRNA.locate_on(target=current_assembly,
fmin=fmin,
fmax=fmax,
strand=strand)
gene.add_mRNA(mRNA)
current_RNA = mRNA
exon_count_by_RNA[transcript_id] = 0
current_CDS_phase = 0
elif args.export_mode == 'cDNA_match':
if current_match is not None and current_match.id != transcript_id:
match.print_as(fh=ofh, source='Cufflinks', format='gff3')
match = things.Match(id=transcript_id,
subclass='cDNA_match',
length=fmax - fmin)
match.locate_on(target=current_assembly,
fmin=fmin,
fmax=fmax,
strand=strand)
current_match = match
elif ftype == 'exon':
exon_number = gff.column_9_value(col9,
'exon_number').replace('"', '')
if args.export_mode == 'model':
exon_count_by_RNA[transcript_id] += 1
cds_id = "{0}.CDS.{1}".format(
current_RNA.id, exon_count_by_RNA[current_RNA.id])
CDS = things.CDS(id=cds_id, parent=current_RNA)
CDS.locate_on(target=current_assembly,
fmin=fmin,
fmax=fmax,
strand=strand,
phase=current_CDS_phase)
current_RNA.add_CDS(CDS)
# calculate the starting phase for the next CDS feature (in case there is one)
current_CDS_phase = 3 - ((
(fmax - fmin) - current_CDS_phase) % 3)
if current_CDS_phase == 3:
current_CDS_phase = 0
exon_id = "{0}.exon.{1}".format(
current_RNA.id, exon_count_by_RNA[current_RNA.id])
exon = things.Exon(id=exon_id, parent=current_RNA)
exon.locate_on(target=current_assembly,
fmin=fmin,
fmax=fmax,
strand=strand)
current_RNA.add_exon(exon)
elif args.export_mode == 'cDNA_match':
mp_id = "{0}.match_part.{1}".format(transcript_id, exon_number)
mp = things.MatchPart(id=mp_id,
parent=current_match,
length=fmax - fmin)
mp.locate_on(target=current_assembly,
fmin=fmin,
fmax=fmax,
strand=strand)
current_match.add_part(mp)
# don't forget to do the last gene, if there were any
if args.export_mode == 'model':
if current_gene is not None:
gene.print_as(fh=ofh, source='GenBank', format='gff3')
elif args.export_mode == 'cDNA_match':
if current_match is not None:
match.print_as(fh=ofh, source='Cufflinks', format='gff3')
def main():
parser = argparse.ArgumentParser(description="Put a description of your script here")
parser.add_argument("-a", "--organism1_annotation", type=str, required=True, help="Annotation GFF for organism 1")
parser.add_argument(
"-p", "--organism1_aat_alignments", type=str, required=True, help="Path to AAT GFF3 (match/match_part)"
)
parser.add_argument(
"-aatdb", "--aat_fasta_db", type=str, required=True, help="Path to FASTA database that was used in AAT"
)
parser.add_argument(
"-b",
"--organism1_blast_alignments",
type=str,
required=True,
help="Path to BLASTp btab file vs.organism 2 proteins",
)
parser.add_argument(
"-be", "--blast_eval_cutoff", type=float, required=False, default=1e-5, help="BLAST e-value cutoff"
)
parser.add_argument(
"-bpi", "--blast_percent_identity_cutoff", type=float, required=False, default=0, help="BLAST %identity cutoff"
)
parser.add_argument(
"-ppc",
"--aat_percent_coverage_cutoff",
type=float,
required=False,
default=0,
help="% coverage of the query protein by the AAT match",
)
parser.add_argument(
"-o", "--output_id_list", type=str, required=False, help="List of IDs from organism1 that passed"
)
args = parser.parse_args()
debugging_transcript = None
## if the output file wasn't passed build one from the other parameters
if args.output_id_list is None:
args.output_id_list = "training_ids.be_{0}.bpi_{1}.ppc_{2}.list".format(
args.blast_eval_cutoff, args.blast_percent_identity_cutoff, args.aat_percent_coverage_cutoff
)
print("INFO: Parsing organism1 annotation")
(assemblies, features) = gff.get_gff3_features(args.organism1_annotation)
print("INFO: Parsing AAT FASTA database")
aat_seqs = utils.fasta_dict_from_file(args.aat_fasta_db)
# keys are assembly IDs, value for each is a list of matches on them
aat_matches = dict()
aat_match_count = 0
current_match = None
## IDs of features in organism 1 which overlap AAT
o1_with_aat = list()
o1_with_o2 = list()
print("INFO: Parsing organism1 AAT protein alignments")
for line in open(args.organism1_aat_alignments):
cols = line.split("\t")
if line.startswith("#") or len(cols) != 9:
continue
assembly_id = cols[0]
# skip this match if there were not predicted genes on the same assembly
if assembly_id not in assemblies:
continue
if assembly_id not in aat_matches:
aat_matches[assembly_id] = list()
fmin = int(cols[3]) - 1
fmax = int(cols[4])
strand = cols[6]
feature_id = gff.column_9_value(cols[8], "ID").replace('"', "")
target = gff.column_9_value(cols[8], "Target")
m = re.search("^(\S+)", target)
if m:
target = m.group(1)
if cols[2] == "nucleotide_to_protein_match":
if current_match is not None:
aat_matches[assembly_id].append(current_match)
aat_match_count += 1
current_match = things.Match(
id=feature_id, target_id=target, subclass="nucleotide_to_protein_match", length=fmax - fmin
)
current_match.locate_on(target=assemblies[assembly_id], fmin=fmin, fmax=fmax, strand=strand)
elif cols[2] == "match_part":
parent_id = gff.column_9_value(cols[8], "Parent").replace('"', "")
match_part = things.MatchPart(id=feature_id, parent=parent_id, length=fmax - fmin)
match_part.locate_on(target=assemblies[assembly_id], fmin=fmin, fmax=fmax, strand=strand)
current_match.add_part(match_part)
print("INFO: Parsed {0} protein alignment chains".format(aat_match_count))
print("INFO: Comparing organism1's mRNAs with AAT match coordinates")
for assembly_id in assemblies:
if assembly_id not in aat_matches:
continue
assembly = assemblies[assembly_id]
for gene in assembly.genes():
for mRNA in gene.mRNAs():
if debugging_transcript is not None:
if mRNA.id == debugging_transcript:
print("DEBUG: processing debugging transcript: {0}".format(mRNA.id))
else:
continue
for aat_match in aat_matches[assembly_id]:
# print("DEBUG: about to call overlap_size_with {0} and {1}, which has {2} segments".format(mRNA.id, aat_match.id, len(aat_match.parts)) )
overlap_size = mRNA.overlap_size_with(aat_match)
if overlap_size is not None:
# print("DEBUG: {0}:({1}) overlaps (size:{2}) {3}:({4})".format(mRNA.id, mRNA.length, overlap_size, aat_match.id, aat_match.length) )
# this shouldn't be possible, but check just in case
if overlap_size > mRNA.length:
raise Exception(
"ERROR: overlap size ({0}) > mRNA length ({1})".format(overlap_size, mRNA.length)
)
if aat_match.target_id not in aat_seqs:
raise Exception(
"ERROR: Found match with target ID ({0}) but didn't find a FASTA entry for it via -aatdb".format(
aat_match.target_id
)
)
# this is a protein length, so x3
match_target_length = len(aat_seqs[aat_match.target_id]["s"]) * 3
(mRNA_percent_coverage, target_percent_coverage) = calculate_fragmented_coverage(
mRNA, aat_match, match_target_length
)
# print("DEBUG: mRNA_percent_coverage:{0}".format(mRNA_percent_coverage) )
# print("DEBUG: match_percent_coverage:{0}".format(target_percent_coverage) )
if (
mRNA_percent_coverage >= args.aat_percent_coverage_cutoff
and target_percent_coverage >= args.aat_percent_coverage_cutoff
):
o1_with_aat.append(mRNA.id)
# print("DEBUG: {0}:({1}) overlaps (size:{2}) {3}:({4}), match target id:{5}, length:{6}".format( \
# mRNA.id, mRNA.length, overlap_size, aat_match.id, aat_match.length, \
# aat_match.target_id, match_target_length) )
# print("\tmRNA % cov: {0}".format(mRNA_percent_coverage))
# print("\ttarget % cov: {0}".format(target_percent_coverage))
break # only need to see if one matched
print("INFO: Found {0} mRNAs in org1 with overlapping fungi AAT coordinates".format(len(o1_with_aat)))
# key=org1_transcript_id, value=org2_transcript_id
top_blast_hits = dict()
print("INFO: parsing BLAST results vs. org2")
for line in open(args.organism1_blast_alignments):
cols = line.split("\t")
if float(cols[19]) > args.blast_eval_cutoff:
continue
if float(cols[10]) < args.blast_percent_identity_cutoff:
continue
# if we survived until here, this one's good.
top_blast_hits[cols[0]] = cols[5]
print("INFO: Comparing overlap between AAT-matched proteins and BLAST ones")
for o1_mRNA_id in o1_with_aat:
if o1_mRNA_id in top_blast_hits:
o1_with_o2.append(o1_mRNA_id)
print(
"INFO: Found {0} mRNAs in org1 with overlapping AAT coordinates and BLAST hit to org2".format(len(o1_with_o2))
)
id_list_fh = open(args.output_id_list, "wt")
for mRNA_id in o1_with_o2:
id_list_fh.write("{0}\n".format(mRNA_id))
last_gene = None
for qry_gene in things:
if qry_gene.id in handled_ids:
continue
## mark this one as handled
handled_ids[qry_gene.id] = 1
nonoverlapping_set.append(qry_gene)
<<<<<<< .mine
current_assembly = assemblies[mol_id]
rfmin = int(cols[3]) - 1
rfmax = int(cols[4])
rstrand = None
feat_id = gff.column_9_value(cols[8], 'ID')
parent_id = gff.column_9_value(cols[8], 'Parent')
parent_feat = None
if parent_id is not None:
if parent_id in features:
parent_feat = features[parent_id]
else:
raise Exception("Error in GFF3: Parent {0} referenced by a child feature before it was defined".format(parent_id) )
#print("Processing feature: ({0})".format(feat_id))
if cols[6] == '-':
strand = -1
elif cols[6] == '+':
strand = 1
def main():
parser = argparse.ArgumentParser( description='Convert PASA GFF file to canonical gene models')
## output file to be written
parser.add_argument('-i', '--input', type=str, required=True, help='Path to a GFF file created by PASA' )
parser.add_argument('-o', '--output', type=str, required=True, help='Path to an output file to be created' )
parser.add_argument('-s', '--source', type=str, required=False, default='PASA', help='Value to use for the 2nd (source) column' )
args = parser.parse_args()
assemblies = dict()
current_assembly = None
gene = None
mRNA = None
gene_fmin = None
gene_fmax = None
gene_strand = None
## Used for tracking the exon count for each gene (for ID purposes)
exon_count_by_mRNA = dict()
fout = open(args.output, mode='wt', encoding='utf-8')
fout.write("##gff-version 3\n")
for line in open(args.input):
cols = line.split("\t")
if len(cols) != 9:
continue
mol_id = cols[0]
feat_type = cols[2]
feat_id = gff.column_9_value(cols[8], 'ID')
# we expect all columns to be cDNA_match
if feat_type != 'cDNA_match':
raise Exception("ERROR: expected all columns to be of type 'cDNA_match' but found a {0}".format(feat_type))
## initialize this assembly if we haven't seen it yet
if mol_id not in assemblies:
assemblies[mol_id] = things.Assembly(id=mol_id)
if gene is None or feat_id != gene.id:
if gene is not None:
# finish the previous one first
mRNA.locate_on( target=current_assembly, fmin=gene_fmin, fmax=gene_fmax, strand=gene_strand )
gene.locate_on( target=current_assembly, fmin=gene_fmin, fmax=gene_fmax, strand=gene_strand )
gene.add_mRNA(mRNA)
current_assembly.add_gene( gene )
gene.print_as(fh=fout, source=args.source, format='gff3')
# now start a new one
gene = things.Gene(id=feat_id)
mRNA = things.mRNA(id="{0}.mRNA".format(feat_id), parent=gene)
exon_count_by_mRNA[mRNA.id] = 0
gene_fmin = int(cols[3]) - 1
gene_fmax = int(cols[4])
gene_strand = cols[6]
current_assembly = assemblies[mol_id]
# each row is a new CDS/exon for the current mRNA
CDS = things.CDS(id="{0}.CDS".format(feat_id), parent=mRNA.id)
# FIX THIS PHASE
CDS.locate_on( target=current_assembly, fmin=int(cols[3]) - 1, fmax=int(cols[4]), strand=cols[6], phase='.' )
mRNA.add_CDS(CDS)
exon_count_by_mRNA[mRNA.id] += 1
exon_id = "{0}.exon{1}".format(mRNA.id, exon_count_by_mRNA[mRNA.id])
exon = things.Exon(id=exon_id, parent=mRNA.id)
exon.locate_on( target=current_assembly, fmin=int(cols[3]) - 1, fmax=int(cols[4]), strand=cols[6] )
mRNA.add_exon(exon)
if int(cols[3]) - 1 < gene_fmin:
gene_fmin = int(cols[3]) - 1
if int(cols[4]) > gene_fmax:
gene_fmax = int(cols[4])
# don't orphan the last one
if gene is not None:
# finish the previous one first
mRNA.locate_on( target=current_assembly, fmin=gene_fmin, fmax=gene_fmax, strand=gene_strand )
gene.locate_on( target=current_assembly, fmin=gene_fmin, fmax=gene_fmax, strand=gene_strand )
gene.add_mRNA(mRNA)
current_assembly.add_gene( gene )
gene.print_as(fh=fout, source=args.source, format='gff3')
def main():
parser = argparse.ArgumentParser(
description='Put a description of your script here')
parser.add_argument('-a',
'--organism1_annotation',
type=str,
required=True,
help='Annotation GFF for organism 1')
parser.add_argument('-p',
'--organism1_aat_alignments',
type=str,
required=True,
help='Path to AAT GFF3 (match/match_part)')
parser.add_argument('-aatdb',
'--aat_fasta_db',
type=str,
required=True,
help='Path to FASTA database that was used in AAT')
parser.add_argument('-b',
'--organism1_blast_alignments',
type=str,
required=True,
help='Path to BLASTp btab file vs.organism 2 proteins')
parser.add_argument('-be',
'--blast_eval_cutoff',
type=float,
required=False,
default=1e-5,
help='BLAST e-value cutoff')
parser.add_argument('-bpi',
'--blast_percent_identity_cutoff',
type=float,
required=False,
default=0,
help='BLAST %identity cutoff')
parser.add_argument(
'-ppc',
'--aat_percent_coverage_cutoff',
type=float,
required=False,
default=0,
help='% coverage of the query protein by the AAT match')
parser.add_argument('-o',
'--output_id_list',
type=str,
required=False,
help='List of IDs from organism1 that passed')
args = parser.parse_args()
debugging_transcript = None
## if the output file wasn't passed build one from the other parameters
if args.output_id_list is None:
args.output_id_list = "training_ids.be_{0}.bpi_{1}.ppc_{2}.list".format(
args.blast_eval_cutoff, args.blast_percent_identity_cutoff,
args.aat_percent_coverage_cutoff)
print("INFO: Parsing organism1 annotation")
(assemblies, features) = gff.get_gff3_features(args.organism1_annotation)
print("INFO: Parsing AAT FASTA database")
aat_seqs = utils.fasta_dict_from_file(args.aat_fasta_db)
# keys are assembly IDs, value for each is a list of matches on them
aat_matches = dict()
aat_match_count = 0
current_match = None
## IDs of features in organism 1 which overlap AAT
o1_with_aat = list()
o1_with_o2 = list()
print("INFO: Parsing organism1 AAT protein alignments")
for line in open(args.organism1_aat_alignments):
cols = line.split("\t")
if line.startswith('#') or len(cols) != 9:
continue
assembly_id = cols[0]
# skip this match if there were not predicted genes on the same assembly
if assembly_id not in assemblies:
continue
if assembly_id not in aat_matches:
aat_matches[assembly_id] = list()
fmin = int(cols[3]) - 1
fmax = int(cols[4])
strand = cols[6]
feature_id = gff.column_9_value(cols[8], 'ID').replace('"', '')
target = gff.column_9_value(cols[8], 'Target')
m = re.search("^(\S+)", target)
if m:
target = m.group(1)
if cols[2] == 'nucleotide_to_protein_match':
if current_match is not None:
aat_matches[assembly_id].append(current_match)
aat_match_count += 1
current_match = things.Match(
id=feature_id,
target_id=target,
subclass='nucleotide_to_protein_match',
length=fmax - fmin)
current_match.locate_on(target=assemblies[assembly_id],
fmin=fmin,
fmax=fmax,
strand=strand)
elif cols[2] == 'match_part':
parent_id = gff.column_9_value(cols[8], 'Parent').replace('"', '')
match_part = things.MatchPart(id=feature_id,
parent=parent_id,
length=fmax - fmin)
match_part.locate_on(target=assemblies[assembly_id],
fmin=fmin,
fmax=fmax,
strand=strand)
current_match.add_part(match_part)
print("INFO: Parsed {0} protein alignment chains".format(aat_match_count))
print("INFO: Comparing organism1's mRNAs with AAT match coordinates")
for assembly_id in assemblies:
if assembly_id not in aat_matches:
continue
assembly = assemblies[assembly_id]
for gene in assembly.genes():
for mRNA in gene.mRNAs():
if debugging_transcript is not None:
if mRNA.id == debugging_transcript:
print("DEBUG: processing debugging transcript: {0}".
format(mRNA.id))
else:
continue
for aat_match in aat_matches[assembly_id]:
#print("DEBUG: about to call overlap_size_with {0} and {1}, which has {2} segments".format(mRNA.id, aat_match.id, len(aat_match.parts)) )
overlap_size = mRNA.overlap_size_with(aat_match)
if overlap_size is not None:
#print("DEBUG: {0}:({1}) overlaps (size:{2}) {3}:({4})".format(mRNA.id, mRNA.length, overlap_size, aat_match.id, aat_match.length) )
# this shouldn't be possible, but check just in case
if overlap_size > mRNA.length:
raise Exception(
"ERROR: overlap size ({0}) > mRNA length ({1})"
.format(overlap_size, mRNA.length))
if aat_match.target_id not in aat_seqs:
raise Exception(
"ERROR: Found match with target ID ({0}) but didn't find a FASTA entry for it via -aatdb"
.format(aat_match.target_id))
# this is a protein length, so x3
match_target_length = len(
aat_seqs[aat_match.target_id]['s']) * 3
(mRNA_percent_coverage, target_percent_coverage
) = calculate_fragmented_coverage(
mRNA, aat_match, match_target_length)
#print("DEBUG: mRNA_percent_coverage:{0}".format(mRNA_percent_coverage) )
#print("DEBUG: match_percent_coverage:{0}".format(target_percent_coverage) )
if mRNA_percent_coverage >= args.aat_percent_coverage_cutoff and target_percent_coverage >= args.aat_percent_coverage_cutoff:
o1_with_aat.append(mRNA.id)
#print("DEBUG: {0}:({1}) overlaps (size:{2}) {3}:({4}), match target id:{5}, length:{6}".format( \
# mRNA.id, mRNA.length, overlap_size, aat_match.id, aat_match.length, \
# aat_match.target_id, match_target_length) )
#print("\tmRNA % cov: {0}".format(mRNA_percent_coverage))
#print("\ttarget % cov: {0}".format(target_percent_coverage))
break # only need to see if one matched
print(
"INFO: Found {0} mRNAs in org1 with overlapping fungi AAT coordinates".
format(len(o1_with_aat)))
# key=org1_transcript_id, value=org2_transcript_id
top_blast_hits = dict()
print("INFO: parsing BLAST results vs. org2")
for line in open(args.organism1_blast_alignments):
cols = line.split("\t")
if float(cols[19]) > args.blast_eval_cutoff:
continue
if float(cols[10]) < args.blast_percent_identity_cutoff:
continue
# if we survived until here, this one's good.
top_blast_hits[cols[0]] = cols[5]
print(
"INFO: Comparing overlap between AAT-matched proteins and BLAST ones")
for o1_mRNA_id in o1_with_aat:
if o1_mRNA_id in top_blast_hits:
o1_with_o2.append(o1_mRNA_id)
print(
"INFO: Found {0} mRNAs in org1 with overlapping AAT coordinates and BLAST hit to org2"
.format(len(o1_with_o2)))
id_list_fh = open(args.output_id_list, 'wt')
for mRNA_id in o1_with_o2:
id_list_fh.write("{0}\n".format(mRNA_id))
def main():
parser = argparse.ArgumentParser( description='Converts glimmerHMM GFF output to GFF3')
# output file to be written
parser.add_argument('-i', '--input_file', type=str, required=True, help='Path to an input file to parse' )
parser.add_argument('-o', '--output_file', type=str, required=True, help='Path to an output file to be created' )
args = parser.parse_args()
fout = open(args.output_file, 'w')
current_gene = None
current_mRNA = None
next_exon_num = defaultdict(int)
for line in open(args.input_file, 'r'):
if line.startswith('#'):
fout.write(line)
continue
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
continue
mol_id = cols[0]
feat_type = cols[2]
feat_fmin = int(cols[3]) - 1
feat_fmax = int(cols[4])
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
if feat_type == 'mRNA':
gene_cols = list(cols)
gene_cols[2] = 'gene'
cols[8] = gff.set_column_9_value(cols[8], 'ID', "{0}.mRNA".format(id))
cols[8] = gff.set_column_9_value(cols[8], 'Name', "{0}.mRNA".format(id))
cols[8] = gff.order_column_9(cols[8])
# print the gene and mRNA
fout.write( "{0}\n".format("\t".join(gene_cols)) )
fout.write( "{0}\n".format("\t".join(cols)) )
elif feat_type == 'CDS':
exon_cols = list(cols)
cols[8] = gff.set_column_9_value(cols[8], 'ID', "{0}.cds".format(parent))
cols[8] = gff.set_column_9_value(cols[8], 'Name', "{0}.cds".format(parent))
cols[8] = gff.set_column_9_value(cols[8], 'Parent', "{0}.mRNA".format(parent))
cols[8] = gff.order_column_9(cols[8])
exon_id = "{0}.exon.{1}".format(parent, next_exon_num[parent] )
next_exon_num[parent] += 1
exon_cols[2] = 'exon'
exon_cols[7] = '.'
exon_cols[8] = gff.set_column_9_value(exon_cols[8], 'ID', exon_id)
exon_cols[8] = gff.set_column_9_value(exon_cols[8], 'Name', exon_id)
exon_cols[8] = gff.set_column_9_value(exon_cols[8], 'Parent', "{0}.mRNA".format(parent))
exon_cols[8] = gff.order_column_9(exon_cols[8])
fout.write( "{0}\n".format("\t".join(exon_cols)) )
fout.write( "{0}\n".format("\t".join(cols)) )
def main():
parser = argparse.ArgumentParser(
description='Convert GFF output from Prodigal into GFF3 format')
## output file to be written
parser.add_argument('-i',
'--input',
type=str,
required=True,
help='Path to a GFF file created by Prodigal')
parser.add_argument('-o',
'--output',
type=str,
required=True,
help='Path to an output file to be created')
args = parser.parse_args()
assemblies = dict()
current_assembly = None
gene = None
mRNAs = dict()
in_sequence = False
current_sequence = None
current_gene_comment_lines = list()
## Used for tracking the exon count for each gene (for ID purposes)
exon_count_by_mRNA = dict()
fout = open(args.output, mode='wt', encoding='utf-8')
fout.write("##gff-version 3\n")
for line in open(args.input):
if line.startswith("#"):
pass
else:
##
gene = None
mRNAs = dict()
in_sequence = False
current_sequence = None
current_gene_comment_lines = list()
##
cols = line.split("\t")
if len(cols) != 9:
continue
mol_id = cols[0]
feat_type = cols[2]
feat_id = gff.column_9_value(cols[8], 'ID')
## initialize this assembly if we haven't seen it yet
if mol_id not in assemblies:
assemblies[mol_id] = things.Assembly(id=mol_id)
current_assembly = assemblies[mol_id]
if feat_type == "CDS":
# gene
gene = things.Gene(id=feat_id)
gene.locate_on(target=current_assembly,
fmin=int(cols[3]) - 1,
fmax=int(cols[4]),
strand=cols[6])
# mRNA
mRNA = things.mRNA(id=feat_id + '.t1', parent=gene)
mRNA.locate_on(target=current_assembly,
fmin=int(cols[3]) - 1,
fmax=int(cols[4]),
strand=cols[6])
gene.add_mRNA(mRNA)
mRNAs[mRNA.id] = mRNA
if feat_id in exon_count_by_mRNA:
raise Exception(
"ERROR: two different mRNAs found with same ID: {0}".
format(feat_id))
else:
exon_count_by_mRNA[feat_id + '.t1'] = 0
# CDS / exons
parent_id = gff.column_9_value(cols[8], 'ID') + '.t1'
## sanity check that we've seen this parent
if parent_id not in mRNAs:
raise Exception(
"ERROR: Found CDS column with parent ({0}) mRNA not yet in the file"
.format(parent_id))
CDS = things.CDS(id=parent_id + '.cds',
parent=mRNAs[parent_id])
CDS.locate_on(target=current_assembly,
fmin=int(cols[3]) - 1,
fmax=int(cols[4]),
strand=cols[6],
phase=int(cols[7]))
mRNA.add_CDS(CDS)
# exons weren't explicitly defined in the input file, so we need to derive new IDs for them
exon_count_by_mRNA[parent_id] += 1
exon_id = "{0}.exon{1}".format(parent_id,
exon_count_by_mRNA[parent_id])
exon = things.Exon(id=exon_id, parent=mRNAs[parent_id])
exon.locate_on(target=current_assembly,
fmin=int(cols[3]) - 1,
fmax=int(cols[4]),
strand=cols[6])
mRNA.add_exon(exon)
##
gene.print_as(fh=fout, source='Prodigal_v2.6.3', format='gff3')
def main():
flawed_gff_file = 'canonical.flawed.gff3'
ilri_gff = 'Theileria-all-Theileria1_ourids.gff'
source = 'GenBank'
out_gff = 'canonical.corrected.gff3'
fout = open(out_gff, mode='wt', encoding='utf-8')
fout.write("##gff-version 3\n")
(assemblies, features) = gff.get_gff3_features(flawed_gff_file)
print("INFO: loaded {0} assemblies and {1} features".format(
len(assemblies), len(features)))
polypeptides = dict()
for line in open(ilri_gff):
cols = line.split("\t")
if len(cols) != 9 or cols[2] != 'polypeptide':
continue
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
polypeptides[parent] = things.Polypeptide(id=id, parent=parent)
polypeptides[parent].locate_on(target=assemblies[cols[0]],
fmin=int(cols[3]) - 1,
fmax=int(cols[4]),
strand=cols[6])
print("DEBUG: loaded {0} polypeptides from ILRI file".format(
len(polypeptides)))
for assembly_id in assemblies:
for gene in assemblies[assembly_id].genes():
for mRNA in gene.mRNAs():
if mRNA.id not in polypeptides:
print(
"DEBUG: {0} not found as a parent to any polypeptide".
format(mRNA.id))
else:
polypeptide = polypeptides[mRNA.id]
# pull this outside of the iteration since iterating might delete some
CDSs = mRNA.CDSs()
for CDS in CDSs:
keep = True
if CDS < polypeptide:
mRNA.delete_CDS(CDS)
elif CDS <= polypeptide:
CDS.location().fmin = polypeptide.location().fmin
if CDS > polypeptide:
mRNA.delete_CDS(CDS)
elif CDS >= polypeptide:
CDS.location().fmax = polypeptide.location().fmax
#print("WARN: found a CDS {0}:{1}-{2} outside the range of the polypeptide {3}:{4}-{5}".format( \
# CDS.id, CDS.locations[0].fmin, CDS.locations[0].fmax, \
# polypeptide.id, polypeptide.locations[0].fmin, polypeptide.locations[0].fmax))
gene.print_as(fh=fout, source=source, format='gff3')
def main():
parser = argparse.ArgumentParser(
'Filter the genes of a GFF3 file by mRNA child IDs')
## output file to be written
parser.add_argument('-i',
'--input_gff3',
type=str,
required=True,
help='GFF3 file of source molecules')
parser.add_argument('-l',
'--id_list',
type=str,
required=True,
help='List file of mRNA IDs to keep')
parser.add_argument('-o',
'--output_file',
type=str,
required=False,
help='Optional output file path (else STDOUT)')
args = parser.parse_args()
## output will either be a file or STDOUT
fout = sys.stdout
if args.output_file is not None:
fout = open(args.output_file, 'wt')
ids_to_keep = list()
for line in open(args.id_list):
line = line.rstrip()
if len(line) > 2:
ids_to_keep.append(line)
fout.write("##gff-version 3\n")
current_gene_lines = list()
current_gene_id = None
keep = False
for line in open(args.input_gff3):
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
continue
# grab the ID and Parent columns if any
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
type = cols[2]
if type == 'gene':
# purge the current gene, if any
if len(current_gene_lines) > 1:
for li in current_gene_lines:
fout.write("{0}\n".format(li))
# reset
current_gene_lines = list()
current_gene_lines.append(line)
current_gene_id = id
else:
if type == 'mRNA':
if id in ids_to_keep:
keep = True
else:
keep = False
if keep == True:
current_gene_lines.append(line)
def main():
parser = argparse.ArgumentParser( description='Adds locus tag identifiers to GFF3 features')
## output file to be written
parser.add_argument('-i', '--input_file', type=str, required=True, help='TA file of source molecules' )
parser.add_argument('-o', '--output_file', type=str, required=False, help='Optional output file path (else STDOUT)' )
parser.add_argument('-p', '--prefix', type=str, required=True, help='The prefix portion of IDs to be generated')
parser.add_argument('-a', '--padding', type=int, required=True, help='Specify the minimum with to reserve for the numeric portion of the IDs. Smaller numbers will be zero-padded.' )
parser.add_argument('-n', '--interval', type=int, required=False, default=1, help='Interval between generated identifiers' )
parser.add_argument('-s', '--starting_id', type=int, required=False, default=0, help='Initial numeric portion of IDs to be generated (do not zero-pad)' )
parser.add_argument('-d', '--id_file', type=str, required=False, help='Pass a 2-column file of IDs to retain (in case you have mapped genes, for example)')
parser.add_argument('-m', '--molecule_map', type=str, required=False, help='Pass a 2-column file of molecule->token identifiers (see documentation)')
parser.add_argument('-c', '--custom', type=str, required=False, help='For custom parsing steps. Most should ignore this.')
args = parser.parse_args()
check_arguments(args)
# used to store locus_tags associated with each gene (so children can inherit)
gene_loci = dict()
next_id = args.starting_id
last_molecule = None
id_mapping = parse_mapping_file( args.id_file )
mol_mapping = parse_mapping_file( args.molecule_map )
loci_assigned = list()
## if using Joana's custom options, check assumptions
if args.custom == 'joana':
if args.molecule_map is None or args.id_file is None:
raise Exception("ERROR: Expected --molecule_map and --id_file options when using --custom=joana")
else:
## need to process the ID map to reformat IDs
for id in id_mapping:
# TP05_0002 -> TpMuguga_05g00002
m = re.match('TP(\d\d)_(\d+)', id_mapping[id])
if m:
id_mapping[id] = "{0}_{1}g0{2}".format(args.prefix, m.group(1), m.group(2) )
elif args.custom == 'bmicroti':
microti_map = { 'I':'01', 'II':'02', 'III':'03', 'IV':'04' }
if args.molecule_map is None or args.id_file is None:
raise Exception("ERROR: Expected --molecule_map and --id_file options when using --custom=bmicroti")
else:
for id in id_mapping:
m = re.match('BBM_(\D+)(\d+)', id_mapping[id])
if m:
print("Changing id from {0} to ".format(id))
id_mapping[id] = "{0}_{1}g{2}".format(args.prefix, microti_map[m.group(1)], m.group(2) )
print(id_mapping[id])
else:
raise Exception("ERROR: id ({0}) didn't match expected convention.".format(id_mapping[id]))
## output will either be a file or STDOUT
fout = sys.stdout
if args.output_file is not None:
fout = open(args.output_file, 'wt')
last_number_portion_assigned = 0
for line in open(args.input_file):
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
fout.write(line + "\n")
continue
if last_molecule is None or (args.molecule_map is not None and mol_mapping[cols[0]] != mol_mapping[last_molecule]):
print("Found molecule {0}, resetting id counter from {1}".format(cols[0], next_id) )
next_id = args.starting_id
last_molecule = cols[0]
# grab the ID column if any
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
type = cols[2]
if type == 'gene':
while True:
if id in id_mapping:
locus_id = id_mapping[id]
else:
if args.molecule_map is None:
locus_id = "{0}_{1}".format(args.prefix, str(next_id).zfill(args.padding))
else:
if cols[0] in mol_mapping:
if args.custom == 'bmicroti':
locus_id = "{0}_{2}g{1}".format(args.prefix, str(int(last_number_portion_assigned) + 1).zfill(args.padding), mol_mapping[cols[0]])
else:
locus_id = "{0}_{2}g{1}".format(args.prefix, str(next_id).zfill(args.padding), mol_mapping[cols[0]])
else:
raise Exception("ERROR: --molecule_map passed but {0} wasn't found in it.".format(cols[0]) )
next_id += args.interval
cols[8] = gff.set_column_9_value(cols[8], 'locus_tag', locus_id)
## make sure this wasn't generated already (possibly conflict between --id_file and an
# auto-generated ID?
if locus_id not in loci_assigned:
break
else:
print("DEBUG: Duplicate ID assigned ({0}), trying again.".format(locus_id) )
loci_assigned.append(locus_id)
gene_loci[id] = locus_id
m = re.search(r"(\d+)$", locus_id)
if m:
last_number_portion_assigned = m.group(1)
elif type.endswith('RNA'):
if parent in gene_loci:
cols[8] = gff.set_column_9_value(cols[8], 'locus_tag', gene_loci[parent])
else:
raise Exception("ERROR: found RNA {0} whose parent {1} wasn't found yet".format(id, parent))
fout.write("\t".join(cols) + "\n")
def main():
parser = argparse.ArgumentParser( description='A GTF -> GFF3 conversion script for StringTie output')
## output file to be written
parser.add_argument('-i', '--input_file', type=str, required=True, help='Path to an input GTF file' )
parser.add_argument('-o', '--output_file', type=str, required=False, help='Path to an output GFF file to be created' )
args = parser.parse_args()
## output will either be a file or STDOUT
ofh = sys.stdout
if args.output_file is not None:
ofh = open(args.output_file, 'wt')
ofh.write("##gff-version 3\n")
assemblies = dict()
current_assembly = None
current_gene = None
current_RNA = None
current_match = None
rna_count_by_gene = defaultdict(int)
exon_count_by_RNA = defaultdict(int)
for line in open(args.input_file, "r"):
cols = line.split("\t")
if len(cols) != 9:
print("SKIPPING: {0}".format(line))
continue
mol_id = cols[0]
if mol_id not in assemblies:
assemblies[mol_id] = things.Assembly(id=mol_id)
current_assembly = assemblies[mol_id]
ftype = cols[2]
fmin = int(cols[3]) - 1
fmax = int(cols[4])
strand = cols[6]
col9 = cols[8]
# this makes it look like GFF column 9 so I can use biocodeutils.column_9_value(str, key)
col9 = col9.replace(' "', '="')
gene_id = gff.column_9_value(col9, 'gene_id').replace('"', '')
transcript_id = gff.column_9_value(col9, 'transcript_id').replace('"', '')
cov = gff.column_9_value(col9, 'cov').replace('"', '')
if ftype == 'transcript':
if current_gene is not None and current_gene.id != gene_id:
gene.print_as(fh=ofh, source='StringTie', format='gff3')
if current_gene is None or current_gene.id != gene_id:
gene = things.Gene(id=gene_id)
gene.locate_on( target=current_assembly, fmin=fmin, fmax=fmax, strand=strand )
current_gene = gene
mRNA = things.mRNA(id=transcript_id, parent=current_gene)
mRNA.locate_on( target=current_assembly, fmin=fmin, fmax=fmax, strand=strand )
gene.add_mRNA(mRNA)
current_RNA = mRNA
exon_count_by_RNA[transcript_id] = 0
current_CDS_phase = 0
elif ftype == 'exon':
exon_number = gff.column_9_value(col9, 'exon_number').replace('"', '')
exon_count_by_RNA[transcript_id] += 1
cds_id = "{0}.CDS.{1}".format( current_RNA.id, exon_count_by_RNA[current_RNA.id] )
CDS = things.CDS(id=cds_id, parent=current_RNA)
CDS.locate_on( target=current_assembly, fmin=fmin, fmax=fmax, strand=strand, phase=current_CDS_phase )
current_RNA.add_CDS(CDS)
# calculate the starting phase for the next CDS feature (in case there is one)
current_CDS_phase = 3 - (((fmax - fmin) - current_CDS_phase) % 3)
if current_CDS_phase == 3:
current_CDS_phase = 0
exon_id = "{0}.exon.{1}".format( current_RNA.id, exon_count_by_RNA[current_RNA.id] )
exon = things.Exon(id=exon_id, parent=current_RNA)
exon.locate_on( target=current_assembly, fmin=fmin, fmax=fmax, strand=strand )
current_RNA.add_exon(exon)
# don't forget to do the last gene, if there were any
if current_gene is not None:
gene.print_as(fh=ofh, source='StringTie', format='gff3')
def main():
parser = argparse.ArgumentParser(
description='Converts glimmerHMM GFF output to GFF3')
# output file to be written
parser.add_argument('-i',
'--input_file',
type=str,
required=True,
help='Path to an input file to parse')
parser.add_argument('-o',
'--output_file',
type=str,
required=True,
help='Path to an output file to be created')
args = parser.parse_args()
fout = open(args.output_file, 'w')
current_gene = None
current_mRNA = None
next_exon_num = defaultdict(int)
for line in open(args.input_file, 'r'):
if line.startswith('#'):
fout.write(line)
continue
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
continue
mol_id = cols[0]
feat_type = cols[2]
feat_fmin = int(cols[3]) - 1
feat_fmax = int(cols[4])
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
if feat_type == 'mRNA':
gene_cols = list(cols)
gene_cols[2] = 'gene'
cols[8] = gff.set_column_9_value(cols[8], 'ID',
"{0}.mRNA".format(id))
cols[8] = gff.set_column_9_value(cols[8], 'Name',
"{0}.mRNA".format(id))
cols[8] = gff.order_column_9(cols[8])
# print the gene and mRNA
fout.write("{0}\n".format("\t".join(gene_cols)))
fout.write("{0}\n".format("\t".join(cols)))
elif feat_type == 'CDS':
exon_cols = list(cols)
cols[8] = gff.set_column_9_value(cols[8], 'ID',
"{0}.cds".format(parent))
cols[8] = gff.set_column_9_value(cols[8], 'Name',
"{0}.cds".format(parent))
cols[8] = gff.set_column_9_value(cols[8], 'Parent',
"{0}.mRNA".format(parent))
cols[8] = gff.order_column_9(cols[8])
exon_id = "{0}.exon.{1}".format(parent, next_exon_num[parent])
next_exon_num[parent] += 1
exon_cols[2] = 'exon'
exon_cols[7] = '.'
exon_cols[8] = gff.set_column_9_value(exon_cols[8], 'ID', exon_id)
exon_cols[8] = gff.set_column_9_value(exon_cols[8], 'Name',
exon_id)
exon_cols[8] = gff.set_column_9_value(exon_cols[8], 'Parent',
"{0}.mRNA".format(parent))
exon_cols[8] = gff.order_column_9(exon_cols[8])
fout.write("{0}\n".format("\t".join(exon_cols)))
fout.write("{0}\n".format("\t".join(cols)))
def main():
parser = argparse.ArgumentParser( description='Removes orphaned features in a GFF3 file')
## output file to be written
parser.add_argument('-i', '--input', type=str, required=True, help='Path to the input GFF3 file' )
parser.add_argument('-o', '--output', type=str, required=True, help='Output GFF3 file to write' )
#parser.add_argument('-t', '--type', type=str, required=False, help='Type of features to remove' )
args = parser.parse_args()
# going to try saving memory by tracking line numbers instead of storing all of it
# true means keep the line, false means to omit it
# doing tracking this way since it's technically legal for a feature to have no identifier at all.
lines = list()
parents = dict()
current_line_num = -1
infile = open(args.input)
for line in infile:
current_line_num += 1
if line.startswith('#'):
lines.append(True)
continue
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
lines.append(True)
continue
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
if parent is None:
# this might be overwritten later
lines.append(False)
if id is not None:
if parent not in parents:
parents[parent] = False
else:
lines.append(True)
parents[parent] = True
infile.seek(0)
current_line_num = -1
outfh = open(args.output, 'wt')
for line in infile:
current_line_num += 1
if lines[current_line_num] == True:
outfh.write(line)
else:
line = line.rstrip()
cols = line.split("\t")
if len(cols) == 9:
id = gff.column_9_value(cols[8], 'ID')
if id is not None and id in parents and parents[id] == True:
outfh.write("{0}\n".format(line))
else:
print("WARN: removing this line: {0}".format(line))
def main():
parser = argparse.ArgumentParser(
description='Adds locus tag identifiers to GFF3 features')
## output file to be written
parser.add_argument('-i',
'--input_file',
type=str,
required=True,
help='TA file of source molecules')
parser.add_argument('-o',
'--output_file',
type=str,
required=False,
help='Optional output file path (else STDOUT)')
parser.add_argument('-p',
'--prefix',
type=str,
required=True,
help='The prefix portion of IDs to be generated')
parser.add_argument(
'-a',
'--padding',
type=int,
required=True,
help=
'Specify the minimum with to reserve for the numeric portion of the IDs. Smaller numbers will be zero-padded.'
)
parser.add_argument('-n',
'--interval',
type=int,
required=False,
default=1,
help='Interval between generated identifiers')
parser.add_argument(
'-s',
'--starting_id',
type=int,
required=False,
default=0,
help='Initial numeric portion of IDs to be generated (do not zero-pad)'
)
parser.add_argument(
'-d',
'--id_file',
type=str,
required=False,
help=
'Pass a 2-column file of IDs to retain (in case you have mapped genes, for example)'
)
parser.add_argument(
'-m',
'--molecule_map',
type=str,
required=False,
help=
'Pass a 2-column file of molecule->token identifiers (see documentation)'
)
parser.add_argument(
'-c',
'--custom',
type=str,
required=False,
help='For custom parsing steps. Most should ignore this.')
args = parser.parse_args()
check_arguments(args)
# used to store locus_tags associated with each gene (so children can inherit)
gene_loci = dict()
next_id = args.starting_id
last_molecule = None
id_mapping = parse_mapping_file(args.id_file)
mol_mapping = parse_mapping_file(args.molecule_map)
loci_assigned = list()
## if using Joana's custom options, check assumptions
if args.custom == 'joana':
if args.molecule_map is None or args.id_file is None:
raise Exception(
"ERROR: Expected --molecule_map and --id_file options when using --custom=joana"
)
else:
## need to process the ID map to reformat IDs
for id in id_mapping:
# TP05_0002 -> TpMuguga_05g00002
m = re.match('TP(\d\d)_(\d+)', id_mapping[id])
if m:
id_mapping[id] = "{0}_{1}g0{2}".format(
args.prefix, m.group(1), m.group(2))
elif args.custom == 'bmicroti':
microti_map = {'I': '01', 'II': '02', 'III': '03', 'IV': '04'}
if args.molecule_map is None or args.id_file is None:
raise Exception(
"ERROR: Expected --molecule_map and --id_file options when using --custom=bmicroti"
)
else:
for id in id_mapping:
m = re.match('BBM_(\D+)(\d+)', id_mapping[id])
if m:
print("Changing id from {0} to ".format(id))
id_mapping[id] = "{0}_{1}g{2}".format(
args.prefix, microti_map[m.group(1)], m.group(2))
print(id_mapping[id])
else:
raise Exception(
"ERROR: id ({0}) didn't match expected convention.".
format(id_mapping[id]))
## output will either be a file or STDOUT
fout = sys.stdout
if args.output_file is not None:
fout = open(args.output_file, 'wt')
last_number_portion_assigned = 0
for line in open(args.input_file):
line = line.rstrip()
cols = line.split("\t")
if len(cols) != 9:
fout.write(line + "\n")
continue
if last_molecule is None or (
args.molecule_map is not None
and mol_mapping[cols[0]] != mol_mapping[last_molecule]):
print("Found molecule {0}, resetting id counter from {1}".format(
cols[0], next_id))
next_id = args.starting_id
last_molecule = cols[0]
# grab the ID column if any
id = gff.column_9_value(cols[8], 'ID')
parent = gff.column_9_value(cols[8], 'Parent')
type = cols[2]
# issue
# 66F4EEF2E3C863C251F831817FF71233
# 7F1917E4D81A959078C9A38E15488BC0
# E22888670919A4A888572155F40F2654
# B9D9CF1F7A8E5A2E1124F0A6C68840DC -> BBM_I00232
# gene before is: 6DE6BCCE69CCDC39994A0940B2ED524A - novel
# errors on: BmicrotiR1_01g00233 -> BBM_I00233
#5800A4110A62E4EAE57AFAD1F8D65CB3 BBM_I00233
if type == 'gene':
while True:
if id in id_mapping:
locus_id = id_mapping[id]
else:
if args.molecule_map is None:
locus_id = "{0}_{1}".format(
args.prefix,
str(next_id).zfill(args.padding))
else:
if cols[0] in mol_mapping:
if args.custom == 'bmicroti':
locus_id = "{0}_{2}g{1}".format(
args.prefix,
str(int(last_number_portion_assigned) +
1).zfill(args.padding),
mol_mapping[cols[0]])
else:
locus_id = "{0}_{2}g{1}".format(
args.prefix,
str(next_id).zfill(args.padding),
mol_mapping[cols[0]])
else:
raise Exception(
"ERROR: --molecule_map passed but {0} wasn't found in it."
.format(cols[0]))
next_id += args.interval
cols[8] = gff.set_column_9_value(cols[8], 'locus_tag',
locus_id)
## make sure this wasn't generated already (possibly conflict between --id_file and an
# auto-generated ID?
if locus_id not in loci_assigned:
break
else:
print("DEBUG: Duplicate ID assigned ({0}), trying again.".
format(locus_id))
loci_assigned.append(locus_id)
gene_loci[id] = locus_id
m = re.search(r"(\d+)$", locus_id)
if m:
last_number_portion_assigned = m.group(1)
elif type.endswith('RNA'):
if parent in gene_loci:
cols[8] = gff.set_column_9_value(cols[8], 'locus_tag',
gene_loci[parent])
else:
raise Exception(
"ERROR: found RNA {0} whose parent {1} wasn't found yet".
format(id, parent))
fout.write("\t".join(cols) + "\n")
def main():
parser = argparse.ArgumentParser( description='Convert native (GTF) or GFF output from Augustus into GFF3 format')
## output file to be written
parser.add_argument('-i', '--input', type=str, required=True, help='Path to a GFF file created by Augustus' )
parser.add_argument('-o', '--output', type=str, required=True, help='Path to an output file to be created' )
args = parser.parse_args()
assemblies = dict()
current_assembly = None
gene = None
mRNAs = dict()
in_sequence = False
current_sequence = None
current_gene_comment_lines = list()
## Used for tracking the exon count for each gene (for ID purposes)
exon_count_by_mRNA = dict()
fout = open(args.output, mode='wt', encoding='utf-8')
fout.write("##gff-version 3\n")
for line in open(args.input):
if line.startswith("#"):
current_gene_comment_lines.append(line)
if line.startswith("# end gene "):
## purge the comments, then write the gene
fout.write( "".join(current_gene_comment_lines) )
gene.print_as(fh=fout, source='AUGUSTUS', format='gff3')
gene = None
mRNAs = dict()
in_sequence = False
current_sequence = None
current_gene_comment_lines = list()
elif line.startswith("# protein sequence = ["):
pass
elif in_sequence is True:
# build 'current_sequence'
pass
else:
cols = line.split("\t")
if len(cols) != 9:
continue
mol_id = cols[0]
feat_type = cols[2]
if feat_type not in ['gene', 'transcript', 'CDS']:
continue
## The output format is GTF by default and (mostly) GFF if the --gff option is used.
# If GTF is detected, let's start by transforming the 9th column into GFF so the
# libraries can use it
# g1 -> ID=g1
# g1.t1 -> ID=g1.t1;Parent=g1
# transcript_id "g1.t1"; gene_id "g1"; -> ID=g1.t1.cds;Parent=g1.t1
m_gene = re.match('(g\d+)', cols[8])
m_transcript = re.match('((g\d+).t\d+)', cols[8])
m_CDS = re.match('transcript_id "(g\d+.t\d+)"; gene_id "g\d+";', cols[8])
# the input can be in GTF or GFF. We need to reformat the 9th column for the GTF entries
if not cols[8].startswith('ID') and not cols[8].startswith('Parent'):
if feat_type == 'gene':
if m_gene:
cols[8] = "ID={0}".format(m_gene.group(1))
else:
raise Exception("ERROR: GTF detected but gene row has bad 9th column format: {0}".format(cols[8]))
elif feat_type == 'transcript':
if m_transcript:
cols[8] = "ID={0};Parent={1}".format(m_transcript.group(1), m_transcript.group(2))
else:
raise Exception("ERROR: GTF detected but transcript row has bad 9th column format: {0}".format(cols[8]))
elif feat_type == 'CDS':
if m_CDS:
cols[8] = "ID={0}.cds;Parent={0}".format(m_CDS.group(1))
else:
raise Exception("ERROR: GTF detected but CDS row has bad 9th column format: {0}".format(cols[8]))
feat_id = gff.column_9_value(cols[8], 'ID')
## initialize this assembly if we haven't seen it yet
if mol_id not in assemblies:
assemblies[mol_id] = things.Assembly(id=mol_id)
current_assembly = assemblies[mol_id]
if feat_type == "gene":
gene = things.Gene(id=feat_id)
gene.locate_on( target=current_assembly, fmin=int(cols[3]) - 1, fmax=int(cols[4]), strand=cols[6] )
elif feat_type == "transcript":
mRNA = things.mRNA(id=feat_id, parent=gene)
mRNA.locate_on( target=current_assembly, fmin=int(cols[3]) - 1, fmax=int(cols[4]), strand=cols[6] )
gene.add_mRNA(mRNA)
mRNAs[mRNA.id] = mRNA
if feat_id in exon_count_by_mRNA:
raise Exception( "ERROR: two different mRNAs found with same ID: {0}".format(feat_id) )
else:
exon_count_by_mRNA[feat_id] = 0
elif feat_type == "CDS":
parent_id = gff.column_9_value(cols[8], 'Parent')
## sanity check that we've seen this parent
if parent_id not in mRNAs:
raise Exception("ERROR: Found CDS column with parent ({0}) mRNA not yet in the file".format(parent_id))
CDS = things.CDS(id=feat_id, parent=mRNAs[parent_id])
CDS.locate_on( target=current_assembly, fmin=int(cols[3]) - 1, fmax=int(cols[4]), strand=cols[6], phase=int(cols[7]) )
mRNA.add_CDS(CDS)
## exons weren't explicitly defined in the input file, so we need to derive new IDs for them
exon_count_by_mRNA[parent_id] += 1
exon_id = "{0}.exon{1}".format(parent_id, exon_count_by_mRNA[parent_id])
exon = things.Exon(id=exon_id, parent=mRNAs[parent_id])
exon.locate_on( target=current_assembly, fmin=int(cols[3]) - 1, fmax=int(cols[4]), strand=cols[6] )
mRNA.add_exon(exon)
