def as2exsj(dstpre, np=7):
ex = UT.read_pandas(dstpre+'.exdf.txt.gz', names=A2.EXDFCOLS)
sj = UT.read_pandas(dstpre+'.sjdf.txt.gz', names=A2.SJDFCOLS)
se = UT.read_pandas(dstpre+'.sedf.txt.gz', names=A2.EXDFCOLS)
paths = UT.read_pandas(dstpre+'.paths.txt.gz', names=A2.PATHCOLS)
#ex.loc[ex['strand'].isin(['.+','.-']),'strand'] = '.'
#sj.loc[sj['strand'].isin(['.+','.-']),'strand'] = '.'
sj['st'] = sj['st']+1
cols = A2.EXDFCOLS
ex = PD.concat([ex[cols],se[cols]],ignore_index=True)
UT.set_info(sj,ex)
UT.set_exon_category(sj, ex)
# find genes (connected components) set '_gidx'
graphpre = dstpre+str(uuid.uuid4())+'_'
prefix = os.path.abspath(graphpre) # need unique prefix for parallel processing
genes = GP.find_genes4(sj,ex,
filepre=prefix,
np=np,
override=False,
separatese=True)
ex.loc[ex['kind']=='3','cat'] = '3'
ex.loc[ex['kind']=='5','cat'] = '5'
UT.write_pandas(ex, dstpre+'.ex.txt.gz', 'h')
UT.write_pandas(sj, dstpre+'.sj.txt.gz', 'h')
ci = UT.chopintervals(ex, dstpre+'.ci.txt.gz')
return sj, ex
def bed2exonsj(bed12, np=4, graphpre=None):
"""Extract exons and junctions from BED12
Args:
bed12: Pandas.DataFrame containing BED12 data
Returns:
sj, ex: Pandas.DataFrames containing junction and exons
"""
esizes = bed12['esizes'].apply(lambda x: N.array([int(y) for y in x.split(',') if y]))
estarts0 = bed12['estarts'].apply(lambda x: N.array([int(y) for y in x.split(',') if y]))
bed12['_estarts'] = bed12['st'] + estarts0
bed12['_eends'] = bed12['_estarts']+esizes
#istarts = eends[:-1]
#iends = estarts[1:]
cols =['chr','st','ed','tname','strand']
def _egen():
for chrom,tname,strand,est,eed in UT.izipcols(bed12,['chr','name','strand','_estarts','_eends']):
if len(est)==1:
yield (chrom,st,ed,tname,0,strand,'s')
else:
if strand=='+':
yield (chrom,est[0],eed[0],tname,0,strand,'5')
for st,ed in izip(est[1:-1],eed[1:-1]):
yield (chrom,st,ed,tname,0,strand,'i')
yield (chrom,est[-1],eed[-1],tname,0,strand,'3')
else: #'-'
yield (chrom,est[0],eed[0],tname,0,strand,'3')
for st,ed in izip(est[1:-1],eed[1:-1]):
yield (chrom,st,ed,tname,0,strand,'i')
yield (chrom,est[-1],eed[-1],tname,0,strand,'5')
def _igen():
for chrom,tname,strand,est,eed in UT.izipcols(bed12,['chr','name','strand','_estarts','_eends']):
#for st,ed in izip(eed[:-1],est[1:]):
for st,ed in izip(eed[:-1],est[1:]):
yield (chrom,st+1,ed,tname,0,strand,'j')
# add 1 to match STAR SJ.tab.out
ex = PD.DataFrame([x for x in _egen()], columns=GGB.BEDCOLS[:6]+['kind'])
ex['locus'] = UT.calc_locus_strand(ex)
ex = ex.groupby('locus').first().reset_index()
sj = PD.DataFrame([x for x in _igen()], columns=GGB.BEDCOLS[:6]+['kind'])
sj['locus'] = UT.calc_locus_strand(sj)
sj = sj.groupby('locus').first().reset_index()
UT.set_info(sj,ex)
UT.set_exon_category(sj, ex)
# find genes (connected components) set '_gidx'
if graphpre is None:
graphpre = './'+str(uuid.uuid4())+'_'
prefix = os.path.abspath(graphpre) # need unique prefix for parallel processing
genes = GP.find_genes4(sj,ex,
filepre=prefix,
np=np,
override=False,
separatese=True)
return sj, ex
def as3exsj(dstpre, minelen=150, np=7):
ex = UT.read_pandas(dstpre+'.exdf.txt.gz', names=A3.EXDFCOLS)
sj = UT.read_pandas(dstpre+'.sjdf.txt.gz', names=A3.SJDFCOLS)
se = UT.read_pandas(dstpre+'.sedf.txt.gz', names=A3.EXDFCOLS)
paths = UT.read_pandas(dstpre+'.paths.txt.gz', names=A3.PATHCOLS)
#ex.loc[ex['strand'].isin(['.+','.-']),'strand'] = '.'
# sj.loc[sj['strand']=='.+','strand'] = '+'
# sj.loc[sj['strand']=='.-','strand'] = '-'
sj['st'] = sj['st']+1
cols = A3.EXDFCOLS
ex = PD.concat([ex[cols],se[cols]],ignore_index=True)
UT.set_info(sj,ex)
UT.set_exon_category(sj, ex)
# find genes (connected components) set '_gidx'
graphpre = dstpre+str(uuid.uuid4())+'_'
prefix = os.path.abspath(graphpre) # need unique prefix for parallel processing
# genes = GP.find_genes4(sj,ex,
# filepre=prefix,
# np=np,
# override=False,
# separatese=True)
genes = GP.find_genes3(sj,ex, # don't use exon overlap as connection
np=np,
override=False)
ex.loc[ex['kind']=='3','cat'] = '3'
ex.loc[ex['kind']=='5','cat'] = '5'
# remove these with elen smaller than minelen
ex['len'] = ex['ed']-ex['st']
exsiz = ex.groupby('_gidx')['len'].sum()
rgidx = exsiz[exsiz<minelen].index.values
LOG.info('minelen filter #ex {0}=>{1}'.format(len(ex), len(ex)-len(rgidx)))
ex2 = ex[~ex['_gidx'].isin(rgidx)]
sj2 = sj[~sj['_gidx'].isin(rgidx)]
# write
UT.write_pandas(ex2, dstpre+'.ex.txt.gz', 'h')
UT.write_pandas(sj2, dstpre+'.sj.txt.gz', 'h')
ci = UT.chopintervals(ex2, dstpre+'.ci.txt.gz')
return sj2, ex2
def test_selectseme(asm):
f = AS.SELECTSEME(asm)
UT.set_exon_category(asm.sj, asm.ae)
f()
def gtf2exonsj(gtf, np=12, graphpre=None):
"""Extract exons and sj from GTF
exon, junction coordinates = zero based (same as BED)
junction start-1 = exon end
junction end = exon start
Args:
gtf: Pandas.DataFrame
Returns:
sj, ex: Pandas.DataFrames for splice junctions and exons
"""
if len(gtf)==0: # edge case
cols = GGB.BEDCOLS[:6]+['locus','_id','cat']
sj = UT.make_empty_df(cols)
ex = UT.make_empty_df(cols)
return sj,ex
exons = gtf[gtf['typ']=='exon'].sort_values(['chr','st','ed'])
exons['_id'] = N.arange(len(exons))
exons.sort_values(['transcript_id','st','ed'],inplace=True)
# 5',3'
ex_s = exons.groupby('transcript_id').size()
tid_s = ex_s[ex_s==1].index
id_m = ex_s[ex_s>1].index
ex_m = exons[exons['transcript_id'].isin(id_m)].copy()
ex_m.sort_values(['transcript_id','st','ed'], inplace=True)
ex_f = ex_m.groupby('transcript_id').first()
ex_l = ex_m.groupby('transcript_id').last()
if5 = list(ex_f[ex_f['strand']=='+']['_id'].values)
if3 = list(ex_f[ex_f['strand']=='-']['_id'].values)
il5 = list(ex_l[ex_l['strand']=='-']['_id'].values)
il3 = list(ex_l[ex_l['strand']=='+']['_id'].values)
exons['kind'] = 'i'
exons.loc[exons['transcript_id'].isin(tid_s),'kind'] = 's'
exons.loc[exons['_id'].isin(if5+il5),'kind'] = '5'
exons.loc[exons['_id'].isin(if3+il3),'kind'] = '3'
# find junctions
def _igen():
for k, g in exons.groupby('transcript_id'):
if len(g)<2:
continue
g = g.sort_values(['st','ed'])
chrom,strand,gid=g.iloc[0][['chr','strand','gene_id']]
ists = g['ed'].values[:-1] + 1
ieds = g['st'].values[1:] - 1
for st,ed in izip(ists,ieds):
# chr,st,ed,name=tid,sc1,strand,gene_id
yield (chrom,st,ed,gid,0,strand)
sj = PD.DataFrame([x for x in _igen()], columns=GGB.BEDCOLS[:6])
sj['locus'] = UT.calc_locus_strand(sj)
sj = sj.groupby('locus').first().reset_index()
#cols = ['chr','st','ed','gene_id','sc1','strand']
#ex = exons #[cols]
exons['locus'] = UT.calc_locus_strand(exons)
ex = exons.groupby(['locus','kind']).first().reset_index() # remove duplicated
ex['name'] = ex['gene_id']
ex['st'] = ex['st'] - 1
# position id == locus converted to number
ex.sort_values(['chr','st','ed'],inplace=True)
ex['_id'] = N.arange(len(ex))
exg = ex.groupby(['chr','st','ed'])[['_id','locus']].first()
exg['_pid'] = N.arange(len(exg)) # position id
cse2pid = dict(zip(exg.index,exg['_pid']))
ex['_pid'] = [cse2pid[tuple(x)] for x in ex[['chr','st','ed']].values]
if len(sj)==0:
ex['_gidx'] = N.arange(len(ex))
return sj, ex
UT.set_info(sj,ex)
UT.set_exon_category(sj, ex)
# find genes (connected components) set '_gidx'
if graphpre is None:
graphpre = './'+str(uuid.uuid4())+'_'
prefix = os.path.abspath(graphpre) # need unique prefix for parallel processing
genes = GP.find_genes4(sj,ex,
filepre=prefix,
np=np,
override=False,
separatese=True)
return sj, ex
def find_genes4(sj,
ae,
filepre,
cachename=None,
np=1,
override=False,
depth=500,
separatese=True):
"""
Adds _gidx column to ae
Connection: 1) by junctions, 2) by overlap in the same strand
Returns genes [set([_id,..]), ...]
"""
if '_id' not in ae.columns:
LOG.info('setting ex _id...')
UT.set_ids(ae)
if '_id' not in sj.columns:
LOG.info('setting sj _id...')
UT.set_ids(sj)
if 'cat' not in ae.columns:
UT.set_exon_category(sj, ae)
if 'a_id' not in ae.columns:
UT.set_ad_info(sj, ae)
### FIND GENES
if cachename and os.path.exists(cachename) and not override:
LOG.info('loading cached genes (connected components)...')
genes = pickle.load(open(cachename, 'rb'))
else:
LOG.info('finding genes (connected components)...')
_sttime = time.time()
if separatese:
me, se = UT.mese(ae)
genes = mcore_allcomponents4(sj, me, filepre, np, depth=depth)
# SE genes
genes += [set([x]) for x in se['_id']]
else:
genes = mcore_allcomponents4(sj, ae, filepre, np, depth=depth)
# version 4 graph: uses overlaps in addition to junctions to connect
# genes = [set([_id's]),...]
if cachename:
UT.makedirs(os.path.dirname(cachename))
pickle.dump(genes, open(cachename, 'wb'))
LOG.info(' time: {0:.3f}s'.format(time.time() - _sttime))
### WRITE EXONS W/ GENE number
LOG.info('assigning gidx...')
_sttime = time.time()
i2g = {} # eid => _gidx
i2gn = {} # eidt => gname
g2gn = {}
i2s = dict(UT.izipcols(ae, ['_id', 'strand'])) # eid => strand
#i2c = dict(UT.izipcols(ae, ['_id','cat'])) # eid => category
s2n = {'+': 'P', '-': 'N', '.': '', '.+': '', '.-': ''}
c2n = {'s': 'S', 'i': 'G', '5': 'G', '3': 'G'}
for i, ids in enumerate(genes):
gid = i + 1
strand = s2n[i2s[list(ids)[0]]]
cat = 'S' if len(ids) == 1 else 'G'
if strand == 'N': # negative strand
gid = -gid
gname = 'J{0}{1}{2}'.format(strand, cat, abs(gid))
g2gn[gid] = gname
for x in ids:
i2g[x] = gid
i2gn[x] = gname
ae['_gidx'] = [i2g[x] for x in ae['_id']]
ae['gname'] = [i2gn[x] for x in ae['_id']]
## set sj _gidx, use acceptor=>_gidx map (exon a_id, sj a_id)
a2g = dict(UT.izipcols(ae, ['a_id', '_gidx']))
d2g = dict(UT.izipcols(ae, ['d_id', '_gidx']))
sj['_gidx'] = [
a2g.get(x, d2g.get(y, 0))
for x, y in UT.izipcols(sj, ['a_id', 'd_id'])
]
sj['gname'] = [g2gn.get(x, '') for x in sj['_gidx']]
# This shouldn't happen
nidx = ae['_gidx'] == 0
if N.sum(nidx) > 0:
LOG.warning(
'###### WARNING!!!!!! exons with no gene assignment:{0}'.format(
N.sum(nidx)))
#ae.loc[nidx, '_gidx'] = N.arange(len(ae),len(ae)+N.sum(nidx))
return genes
