def try_to_combine(self, irandom):
"""
Create a recombination event and write it to disk
<irandom> is used as the seed for the myriad random number calls.
If combine() is called with the same <irandom>, it will find the same event, i.e. it should be a random number, not just a seed
"""
if self.args.debug:
print 'combine (seed %d)' % irandom
numpy.random.seed(irandom)
random.seed(irandom)
reco_event = RecombinationEvent(self.glfo)
self.choose_vdj_combo(reco_event)
self.erode_and_insert(reco_event)
if self.args.debug:
print ' joining eroded seqs'
print ' v: %s' % reco_event.eroded_seqs['v']
print ' insert: %s' % reco_event.insertions['vd']
print ' d: %s' % reco_event.eroded_seqs['d']
print ' insert: %s' % reco_event.insertions['dj']
print ' j: %s' % reco_event.eroded_seqs['j']
reco_event.recombined_seq = reco_event.eroded_seqs[
'v'] + reco_event.insertions['vd'] + reco_event.eroded_seqs[
'd'] + reco_event.insertions['dj'] + reco_event.eroded_seqs['j']
reco_event.set_post_erosion_codon_positions()
# set the original conserved codon words, so we can revert them if they get mutated NOTE we do it here, *after* setting the full recombined sequence, so the germline Vs that don't extend through the cysteine don't screw us over
reco_event.unmutated_codons = {}
for region, codon in utils.conserved_codons[self.args.locus].items():
fpos = reco_event.post_erosion_codon_positions[region]
original_codon = reco_event.recombined_seq[fpos:fpos + 3]
reco_event.unmutated_codons[region] = reco_event.recombined_seq[
fpos:fpos + 3]
# print fpos, original_codon, utils.codon_unmutated(codon, reco_event.recombined_seq, fpos)
codons_ok = utils.both_codons_unmutated(
self.glfo['locus'],
reco_event.recombined_seq,
reco_event.post_erosion_codon_positions,
extra_str=' ',
debug=self.args.debug)
if not codons_ok:
if self.args.rearrange_from_scratch and self.args.generate_germline_set: # if you let it try more than once, it screws up the desired allele prevalence ratios
raise Exception('arg')
return False
in_frame = utils.in_frame(
reco_event.recombined_seq, reco_event.post_erosion_codon_positions,
'', reco_event.effective_erosions['v_5p']
) # NOTE empty string is the fv insertion, which is hard coded to zero in event.py. I no longer recall the details of that decision, but I have a large amount of confidence that it's more sensible than it looks
if self.args.rearrange_from_scratch and not in_frame:
raise Exception(
'arg 2'
) # if you let it try more than once, it screws up the desired allele prevalence ratios
return False
self.add_mutants(reco_event, irandom)
reco_event.write_event(self.outfname, reco_event.line)
return True
def try_to_combine(self, irandom):
"""
Create a recombination event and write it to disk
<irandom> is used as the seed for the myriad random number calls.
If combine() is called with the same <irandom>, it will find the same event, i.e. it should be a random number, not just a seed
"""
if self.args.debug:
print 'combine (seed %d)' % irandom
numpy.random.seed(irandom)
random.seed(irandom)
reco_event = RecombinationEvent(self.glfo)
self.choose_vdj_combo(reco_event)
self.erode_and_insert(reco_event)
if self.args.debug:
print ' joining eroded seqs'
print ' v: %s' % reco_event.eroded_seqs['v']
print ' insert: %s' % reco_event.insertions['vd']
print ' d: %s' % reco_event.eroded_seqs['d']
print ' insert: %s' % reco_event.insertions['dj']
print ' j: %s' % reco_event.eroded_seqs['j']
reco_event.recombined_seq = reco_event.eroded_seqs['v'] + reco_event.insertions['vd'] + reco_event.eroded_seqs['d'] + reco_event.insertions['dj'] + reco_event.eroded_seqs['j']
reco_event.set_final_codon_positions()
# set the original conserved codon words, so we can revert them if they get mutated NOTE we do it here, *after* setting the full recombined sequence, so the germline Vs that don't extend through the cysteine don't screw us over
reco_event.unmutated_codons = {}
for region, codon in utils.conserved_codons[self.args.chain].items():
fpos = reco_event.final_codon_positions[region]
original_codon = reco_event.recombined_seq[fpos : fpos + 3]
reco_event.unmutated_codons[region] = reco_event.recombined_seq[fpos : fpos + 3]
# print fpos, original_codon, utils.codon_ok(codon, reco_event.recombined_seq, fpos)
codons_ok = utils.both_codons_ok(self.glfo['chain'], reco_event.recombined_seq, reco_event.final_codon_positions, extra_str=' ', debug=self.args.debug)
if not codons_ok:
if self.args.rearrange_from_scratch and self.args.generate_germline_set: # if you let it try more than once, it screws up the desired allele prevalence ratios
raise Exception('arg')
return False
in_frame = reco_event.cdr3_length % 3 == 0
if self.args.rearrange_from_scratch and not in_frame:
raise Exception('arg 2') # if you let it try more than once, it screws up the desired allele prevalence ratios
return False
self.add_mutants(reco_event, irandom) # toss a bunch of clones: add point mutations
if self.args.debug:
reco_event.print_event()
# write output to csv
reco_event.write_event(self.outfname, irandom=irandom)
return True
def try_to_combine(self, irandom):
"""
Create a recombination event and write it to disk
<irandom> is used as the seed for the myriad random number calls.
If combine() is called with the same <irandom>, it will find the same event, i.e. it should be a random number, not just a seed
"""
if self.args.debug:
print 'combine (seed %d)' % irandom
numpy.random.seed(irandom)
random.seed(irandom)
reco_event = RecombinationEvent(self.glfo)
self.choose_vdj_combo(reco_event)
self.erode_and_insert(reco_event)
if self.args.debug:
print ' joining eroded seqs'
print ' v: %s' % reco_event.eroded_seqs['v']
print ' insert: %s' % reco_event.insertions['vd']
print ' d: %s' % reco_event.eroded_seqs['d']
print ' insert: %s' % reco_event.insertions['dj']
print ' j: %s' % reco_event.eroded_seqs['j']
reco_event.recombined_seq = reco_event.eroded_seqs['v'] + reco_event.insertions['vd'] + reco_event.eroded_seqs['d'] + reco_event.insertions['dj'] + reco_event.eroded_seqs['j']
codons_ok = reco_event.set_final_cyst_tryp_positions(debug=self.args.debug)
if not codons_ok:
return False
self.add_mutants(reco_event, irandom) # toss a bunch of clones: add point mutations
if self.args.debug:
reco_event.print_event()
# write output to csv
reco_event.write_event(self.outfname, irandom=irandom)
return True
def combine(self, irandom):
"""
Create a recombination event and write it to disk
<irandom> is used as the seed for the myriad random number calls.
If combine() is called with the same <irandom>, it will find the same event, i.e. it should be a random number, not just a seed
"""
if self.args.debug:
print 'combine (seed %d)' % irandom
numpy.random.seed(irandom)
random.seed(irandom)
reco_event = RecombinationEvent(self.all_seqs)
self.choose_vdj_combo(reco_event)
self.erode_and_insert(reco_event)
if self.args.debug:
print ' joining eroded seqs'
print ' v: %s' % reco_event.eroded_seqs['v']
print ' insert: %s' % reco_event.insertions['vd']
print ' d: %s' % reco_event.eroded_seqs['d']
print ' insert: %s' % reco_event.insertions['dj']
print ' j: %s' % reco_event.eroded_seqs['j']
reco_event.recombined_seq = reco_event.eroded_seqs[
'v'] + reco_event.insertions['vd'] + reco_event.eroded_seqs[
'd'] + reco_event.insertions['dj'] + reco_event.eroded_seqs['j']
try:
reco_event.set_final_cyst_tryp_positions(
total_length_from_right=self.total_length_from_right,
debug=self.args.debug)
except AssertionError:
print 'ERROR bad conserved codos, what the hell?'
return False
if self.args.naivety == 'M':
self.add_mutants(
reco_event,
irandom) # toss a bunch of clones: add point mutations
else:
reco_event.final_seqs.append(reco_event.recombined_seq)
if self.args.debug:
reco_event.print_event(self.total_length_from_right)
# write output to csv
reco_event.write_event(self.outfname,
self.total_length_from_right,
irandom=irandom)
return True
def combine(self, irandom):
"""
Create a recombination event and write it to disk
<irandom> is used as the seed for the myriad random number calls.
If combine() is called with the same <irandom>, it will find the same event, i.e. it should be a random number, not just a seed
"""
if self.args.debug:
print 'combine (seed %d)' % irandom
numpy.random.seed(irandom)
random.seed(irandom)
reco_event = RecombinationEvent(self.all_seqs)
self.choose_vdj_combo(reco_event)
self.erode_and_insert(reco_event)
if self.args.debug:
print ' joining eroded seqs'
print ' v: %s' % reco_event.eroded_seqs['v']
print ' insert: %s' % reco_event.insertions['vd']
print ' d: %s' % reco_event.eroded_seqs['d']
print ' insert: %s' % reco_event.insertions['dj']
print ' j: %s' % reco_event.eroded_seqs['j']
reco_event.recombined_seq = reco_event.eroded_seqs['v'] + reco_event.insertions['vd'] + reco_event.eroded_seqs['d'] + reco_event.insertions['dj'] + reco_event.eroded_seqs['j']
try:
reco_event.set_final_cyst_tryp_positions(total_length_from_right=self.total_length_from_right, debug=self.args.debug)
except AssertionError:
print 'ERROR bad conserved codos, what the hell?'
return False
if self.args.naivety == 'M':
self.add_mutants(reco_event, irandom) # toss a bunch of clones: add point mutations
else:
reco_event.final_seqs.append(reco_event.recombined_seq)
if self.args.debug:
reco_event.print_event(self.total_length_from_right)
# write output to csv
reco_event.write_event(self.outfname, self.total_length_from_right, irandom=irandom)
return True
def parse_bcr_phylo_output(glfo, naive_line, outdir, ievent):
seqfos = utils.read_fastx(bcr_phylo_fasta_fname(
outdir)) # output mutated sequences from bcr-phylo
assert len(
naive_line['unique_ids']
) == 1 # enforces that we ran naive-only, 1-leaf partis simulation above
assert not indelutils.has_indels(
naive_line['indelfos'][0]) # would have to handle this below
if args.debug:
utils.print_reco_event(naive_line)
reco_info = collections.OrderedDict()
for sfo in seqfos:
mline = copy.deepcopy(naive_line)
utils.remove_all_implicit_info(mline)
del mline['tree']
mline['unique_ids'] = [sfo['name']]
mline['seqs'] = [
sfo['seq']
] # it's really important to set both the seqs (since they're both already in there from the naive line)
mline['input_seqs'] = [
sfo['seq']
] # it's really important to set both the seqs (since they're both already in there from the naive line)
mline['duplicates'] = [[]]
reco_info[sfo['name']] = mline
utils.add_implicit_info(glfo, mline)
final_line = utils.synthesize_multi_seq_line_from_reco_info(
[sfo['name'] for sfo in seqfos], reco_info)
if args.debug:
utils.print_reco_event(final_line)
# extract kd values from pickle file (use a separate script since it requires ete/anaconda to read)
if args.stype == 'selection':
cmd = './bin/read-bcr-phylo-trees.py --pickle-tree-file %s/%s_lineage_tree.p --kdfile %s/kd-vals.csv --newick-tree-file %s/simu.nwk' % (
outdir, args.extrastr, outdir, outdir)
utils.run_ete_script(cmd, ete_path)
nodefo = {}
with open('%s/kd-vals.csv' % outdir) as kdfile:
reader = csv.DictReader(kdfile)
for line in reader:
nodefo[line['uid']] = {
'kd': float(line['kd']),
'relative_kd': float(line['relative_kd']),
'lambda': line.get('lambda', None),
'target_index': int(line['target_index']),
}
if len(
set(nodefo) - set(final_line['unique_ids'])
) > 0: # uids in the kd file but not the <line> (i.e. not in the newick/fasta files) are probably just bcr-phylo discarding internal nodes
print ' in kd file, but missing from final_line (probably just internal nodes that bcr-phylo wrote to the tree without names): %s' % (
set(nodefo) - set(final_line['unique_ids']))
if len(set(final_line['unique_ids']) - set(nodefo)) > 0:
print ' in final_line, but missing from kdvals: %s' % ' '.join(
set(final_line['unique_ids']) - set(nodefo))
final_line['affinities'] = [
1. / nodefo[u]['kd'] for u in final_line['unique_ids']
]
final_line['relative_affinities'] = [
1. / nodefo[u]['relative_kd'] for u in final_line['unique_ids']
]
final_line['lambdas'] = [
nodefo[u]['lambda'] for u in final_line['unique_ids']
]
final_line['nearest_target_indices'] = [
nodefo[u]['target_index'] for u in final_line['unique_ids']
]
tree = treeutils.get_dendro_tree(treefname='%s/simu.nwk' % outdir)
tree.scale_edges(1. / numpy.mean([len(s) for s in final_line['seqs']]))
if args.debug:
print utils.pad_lines(treeutils.get_ascii_tree(dendro_tree=tree),
padwidth=12)
final_line['tree'] = tree.as_string(schema='newick')
tmp_event = RecombinationEvent(
glfo) # I don't want to move the function out of event.py right now
tmp_event.set_reco_id(
final_line, irandom=ievent
) # not sure that setting <irandom> here actually does anything
# get target sequences
target_seqfos = utils.read_fastx('%s/%s_targets.fa' %
(outdir, args.extrastr))
final_line['target_seqs'] = [tfo['seq'] for tfo in target_seqfos]
return final_line
def parse_bcr_phylo_output(glfo, naive_line, outdir, ievent):
seqfos = utils.read_fastx(
'%s/%s.fasta' %
(outdir, args.extrastr)) # output mutated sequences from bcr-phylo
assert len(
naive_line['unique_ids']
) == 1 # enforces that we ran naive-only, 1-leaf partis simulation above
assert not indelutils.has_indels(
naive_line['indelfos'][0]) # would have to handle this below
if args.debug:
utils.print_reco_event(naive_line)
reco_info = collections.OrderedDict()
for sfo in seqfos:
mline = copy.deepcopy(naive_line)
utils.remove_all_implicit_info(mline)
del mline['tree']
mline['unique_ids'] = [sfo['name']]
mline['seqs'] = [
sfo['seq']
] # it's really important to set both the seqs (since they're both already in there from the naive line)
mline['input_seqs'] = [
sfo['seq']
] # it's really important to set both the seqs (since they're both already in there from the naive line)
reco_info[sfo['name']] = mline
utils.add_implicit_info(glfo, mline)
final_line = utils.synthesize_multi_seq_line_from_reco_info(
[sfo['name'] for sfo in seqfos], reco_info)
if args.debug:
utils.print_reco_event(final_line)
# extract kd values from pickle file (use a separate script since it requires ete/anaconda to read)
if args.stype == 'selection':
cmd = 'export PATH=%s:$PATH && xvfb-run -a python ./bin/view-trees.py --pickle-tree-file %s/%s_lineage_tree.p --kdfile %s/kd-vals.csv --newick-tree-file %s/simu.nwk' % (
ete_path, outdir, args.extrastr, outdir, outdir)
utils.simplerun(cmd, shell=True)
kdvals = {}
with open('%s/kd-vals.csv' % outdir) as kdfile:
reader = csv.DictReader(kdfile)
for line in reader:
kdvals[line['uid']] = float(line['kd'])
if len(
set(kdvals) - set(final_line['unique_ids'])
) > 0: # uids in the kd file but not the <line> (i.e. not in the newick/fasta files) are probably just bcr-phylo discarding internal nodes
print ' in kd file, but missing from final_line (probably just internal nodes that bcr-phylo wrote to the tree without names): %s' % (
set(kdvals) - set(final_line['unique_ids']))
if len(set(final_line['unique_ids']) - set(kdvals)) > 0:
print ' in final_line, but missing from kdvals: %s' % ' '.join(
set(final_line['unique_ids']) - set(kdvals))
final_line['affinities'] = [
1. / kdvals[u] for u in final_line['unique_ids']
]
tree = treeutils.get_dendro_tree(treefname='%s/simu.nwk' % outdir)
if args.debug:
print utils.pad_lines(treeutils.get_ascii_tree(dendro_tree=tree),
padwidth=12)
final_line['tree'] = tree.as_string(schema='newick')
tmp_event = RecombinationEvent(
glfo) # I don't want to move the function out of event.py right now
tmp_event.set_reco_id(
final_line, irandom=ievent
) # not sure that setting <irandom> here actually does anything
# get target sequences
target_seqfos = utils.read_fastx('%s/%s_targets.fa' %
(outdir, args.extrastr))
final_line['target_seqs'] = [tfo['seq'] for tfo in target_seqfos]
from Bio.Seq import Seq
final_line['nearest_target_indices'] = []
aa_targets = [Seq(seq).translate() for seq in final_line['target_seqs']]
for mseq in final_line['input_seqs']:
aa_mseq = Seq(mseq).translate()
aa_hdists = [
utils.hamming_distance(aa_t, aa_mseq, amino_acid=True)
for aa_t in aa_targets
]
imin = aa_hdists.index(
min(aa_hdists)
) # NOTE doesn't do anything differently if there's more than one min
final_line['nearest_target_indices'].append(imin)
return final_line
def parse_bcr_phylo_output(glfo, naive_line, outdir, ievent):
seqfos = utils.read_fastx(bcr_phylo_fasta_fname(outdir)) # output mutated sequences from bcr-phylo
assert len(naive_line['unique_ids']) == 1 # enforces that we ran naive-only, 1-leaf partis simulation above
assert not indelutils.has_indels(naive_line['indelfos'][0]) # would have to handle this below
if args.debug:
utils.print_reco_event(naive_line)
reco_info = collections.OrderedDict()
for sfo in seqfos:
mline = copy.deepcopy(naive_line)
utils.remove_all_implicit_info(mline)
del mline['tree']
mline['unique_ids'] = [sfo['name']]
mline['seqs'] = [sfo['seq']] # it's really important to set both the seqs (since they're both already in there from the naive line)
mline['input_seqs'] = [sfo['seq']] # it's really important to set both the seqs (since they're both already in there from the naive line)
mline['duplicates'] = [[]]
reco_info[sfo['name']] = mline
try:
utils.add_implicit_info(glfo, mline)
except: # TODO not sure if I really want to leave this in long term, but it shouldn't hurt anything (it's crashing on unequal naive/mature sequence lengths, and I need this to track down which event it is) UPDATE: yeah it was just because something crashed in the middle of writing a .fa file
print 'implicit info adding failed for ievent %d in %s' % (ievent, outdir)
lines = traceback.format_exception(*sys.exc_info())
print utils.pad_lines(''.join(lines)) # NOTE this will still crash on the next line if implicit info adding failed
final_line = utils.synthesize_multi_seq_line_from_reco_info([sfo['name'] for sfo in seqfos], reco_info)
if args.debug:
utils.print_reco_event(final_line)
# extract kd values from pickle file (use a separate script since it requires ete/anaconda to read)
if args.stype == 'selection':
kdfname, nwkfname = '%s/kd-vals.csv' % outdir, '%s/simu.nwk' % outdir
if not utils.output_exists(args, kdfname, outlabel='kd/nwk conversion', offset=4): # eh, don't really need to check for both kd an nwk file, chances of only one being missing are really small, and it'll just crash when it looks for it a couple lines later
cmd = './bin/read-bcr-phylo-trees.py --pickle-tree-file %s/%s_lineage_tree.p --kdfile %s --newick-tree-file %s' % (outdir, args.extrastr, kdfname, nwkfname)
utils.run_ete_script(cmd, ete_path, debug=args.n_procs==1)
nodefo = {}
with open(kdfname) as kdfile:
reader = csv.DictReader(kdfile)
for line in reader:
nodefo[line['uid']] = {
'kd' : float(line['kd']),
'relative_kd' : float(line['relative_kd']),
'lambda' : line.get('lambda', None),
'target_index' : int(line['target_index']),
}
if len(set(nodefo) - set(final_line['unique_ids'])) > 0: # uids in the kd file but not the <line> (i.e. not in the newick/fasta files) are probably just bcr-phylo discarding internal nodes
print ' in kd file, but missing from final_line (probably just internal nodes that bcr-phylo wrote to the tree without names): %s' % (set(nodefo) - set(final_line['unique_ids']))
if len(set(final_line['unique_ids']) - set(nodefo)) > 0:
print ' in final_line, but missing from kdvals: %s' % ' '.join(set(final_line['unique_ids']) - set(nodefo))
final_line['affinities'] = [1. / nodefo[u]['kd'] for u in final_line['unique_ids']]
final_line['relative_affinities'] = [1. / nodefo[u]['relative_kd'] for u in final_line['unique_ids']]
final_line['lambdas'] = [nodefo[u]['lambda'] for u in final_line['unique_ids']]
final_line['nearest_target_indices'] = [nodefo[u]['target_index'] for u in final_line['unique_ids']]
tree = treeutils.get_dendro_tree(treefname=nwkfname)
tree.scale_edges(1. / numpy.mean([len(s) for s in final_line['seqs']]))
if args.debug:
print utils.pad_lines(treeutils.get_ascii_tree(dendro_tree=tree), padwidth=12)
final_line['tree'] = tree.as_string(schema='newick')
tmp_event = RecombinationEvent(glfo) # I don't want to move the function out of event.py right now
tmp_event.set_reco_id(final_line, irandom=ievent) # not sure that setting <irandom> here actually does anything
# get target sequences
target_seqfos = utils.read_fastx('%s/%s_targets.fa' % (outdir, args.extrastr))
final_line['target_seqs'] = [tfo['seq'] for tfo in target_seqfos]
return final_line
def get_mature_line(sfos,
naive_line,
glfo,
nodefo,
dtree,
target_sfos,
locus=None):
assert len(
naive_line['unique_ids']
) == 1 # enforces that we ran naive-only, 1-leaf partis simulation above
assert not indelutils.has_indels(
naive_line['indelfos'][0]) # would have to handle this below
if args.debug:
utils.print_reco_event(naive_line)
reco_info = collections.OrderedDict()
for sfo in sfos:
mline = utils.get_non_implicit_copy(naive_line)
del mline['tree']
mline['unique_ids'] = [sfo['name']]
mline['seqs'] = [sfo['seq']]
mline['input_seqs'] = [
sfo['seq']
] # it's really important to set both the seqs (since they're both already in there from the naive line)
mline['duplicates'] = [[]]
reco_info[sfo['name']] = mline
try:
utils.add_implicit_info(glfo, mline)
except: # TODO not sure if I really want to leave this in long term, but it shouldn't hurt anything (it's crashing on unequal naive/mature sequence lengths, and I need this to track down which event it is) UPDATE: yeah it was just because something crashed in the middle of writing a .fa file
print 'implicit info adding failed for ievent %d in %s' % (
ievent, outdir)
lines = traceback.format_exception(*sys.exc_info())
print utils.pad_lines(
''.join(lines)
) # NOTE this will still crash on the next line if implicit info adding failed
final_line = utils.synthesize_multi_seq_line_from_reco_info(
[sfo['name'] for sfo in sfos], reco_info)
ftree = copy.deepcopy(dtree)
if locus is not None:
def ltr(u):
return u + '-' + locus
new_nodefo = {}
for u_old in nodefo:
new_nodefo[ltr(u_old)] = nodefo[u_old]
nodefo = new_nodefo
treeutils.translate_labels(ftree,
[(u, ltr(u))
for u in final_line['unique_ids']])
final_line['unique_ids'] = [
ltr(u) for u in final_line['unique_ids']
]
assert len(sfos) == len(final_line['unique_ids'])
for iseq, sfo in enumerate(sfos):
naive_id = naive_line['unique_ids'][0]
assert naive_id.count('-') == 1
bstr = naive_id.replace('-' + locus, '')
pids = final_line['paired-uids'][iseq]
assert len(pids) == 1 and pids[0].find(
bstr
) == 0 and pids[0].count('-') == 1 and pids[0].split(
'-'
)[1] in utils.loci # if uid is xxx-igh, paired id shoud be e.g. xxx-igk
final_line['paired-uids'][iseq] = [
p.replace(bstr, sfo['name']) for p in pids
]
if args.debug:
utils.print_reco_event(final_line)
# extract kd values from pickle file (use a separate script since it requires ete/anaconda to read)
if len(
set(nodefo) - set(final_line['unique_ids'])
) > 0: # uids in the kd file but not the <line> (i.e. not in the newick/fasta files) are probably just bcr-phylo discarding internal nodes
print ' in kd file, but missing from final_line (probably just internal nodes that bcr-phylo wrote to the tree without names): %s' % (
set(nodefo) - set(final_line['unique_ids']))
if len(set(final_line['unique_ids']) - set(nodefo)) > 0:
print ' in final_line, but missing from kdvals: %s' % ' '.join(
set(final_line['unique_ids']) - set(nodefo))
final_line['affinities'] = [
1. / nodefo[u]['kd'] for u in final_line['unique_ids']
]
final_line['relative_affinities'] = [
1. / nodefo[u]['relative_kd'] for u in final_line['unique_ids']
]
final_line['lambdas'] = [
nodefo[u]['lambda'] for u in final_line['unique_ids']
]
final_line['nearest_target_indices'] = [
nodefo[u]['target_index'] for u in final_line['unique_ids']
]
ftree.scale_edges(1. / numpy.mean([len(s)
for s in final_line['seqs']]))
if args.debug:
print utils.pad_lines(treeutils.get_ascii_tree(dendro_tree=ftree),
padwidth=12)
final_line['tree'] = ftree.as_string(schema='newick')
tmp_event = RecombinationEvent(
glfo
) # I don't want to move the function out of event.py right now
tmp_event.set_reco_id(
final_line, irandom=ievent
) # not sure that setting <irandom> here actually does anything
final_line['target_seqs'] = [tfo['seq'] for tfo in target_sfos]
return final_line
