def gba_map_taxa(dbname):
cbdb = compbio.projects.cbdb
dbi = cbdb.getName(dbname)
tgb_dbi = cbdb.getName('tax_gbs')
count = 0
slc_n = 10000
slc_ofs = 0
max_ofs =dbi.S.q(func.max(dbi.Sequence.id)).scalar()
while slc_ofs < max_ofs:
for s in dbi.Session.query(dbi.Sequence).\
filter(dbi.Sequence.id>=slc_ofs).\
filter(dbi.Sequence.id<(slc_ofs+slc_n)):
gbid = s.gb_id
count += 1
if gbid:
taxid = tgb_dbi.Session.query(tgb_dbi.TaxGBJoin).\
filter_by(gbid = gbid).first()
s.source_taxon = taxid.taxid
if count > 100:
dbi.Session.commit()
print 'committing'
count = 0
slc_ofs += slc_n
dbi.Session.commit()
def db_metadata(clade):
'''
Get an ncbi genealogy for a clade - e.g: the minimal ncbi node containing
every terminal of a clade as well as the ncbi nodes at each leaf.
inputs:
clade: <biopython clade>
outputs:
genealogy: the shared ncbi genealogy of every terminal
nodes: ncbi taxnodes for every terminal
'''
tax_dbi = cbdb.getName('taxdmp')
tgb_dbi = cbdb.getName('tax_gbs')
gba_dbi = cbdb.getName('gb_acc_idjoin')
#terminal_gbaccs = map(lambda x: clade_gbacc(x), clade.get_terminals())
for idx, t in enumerate(clade.get_terminals()):
try:
t.m['id'] = idx
t.m['gbacc'] = clade_gbacc(t)
t.m['gbid'] = gba_dbi.Session.query(gba_dbi.GBAcc).\
filter_by(accession = t.m['gbacc']).one().gbid
taxid = tgb_dbi.Session.query(tgb_dbi.TaxGBJoin).\
filter_by(gbid = t.m['gbid']).one().taxid
t.m['taxid'] = taxid
except: pass
max_idx = idx
for idx, t in enumerate(clade.get_nonterminals()):
t.m['id'] = max_idx + idx
def gba_map_gbids(dbname):
cbdb = compbio.projects.cbdb
dbi = cbdb.getName(dbname)
count = 0
fail_count =0
sxs_count =0
gbacc_dbi = cbdb.getName('gb_acc_idjoin')
slc_n = 10000
slc_ofs = 0
max_ofs =dbi.S.q(func.max(dbi.Sequence.id)).scalar()
while slc_ofs < max_ofs:
for s in dbi.Session.query(dbi.Sequence).\
filter(dbi.Sequence.id>=slc_ofs).\
filter(dbi.Sequence.id<(slc_ofs+slc_n)):
try:
gbid = gbacc_dbi.Session.query(gbacc_dbi.GBAcc).\
filter_by(accession = s.gb_accession).first().gbid
s.gb_id = gbid
sxs_count += 1
except:
print 'failed!'
fail_count +=1
count += 1
if count > 100:
dbi.Session.commit()
count = 0
print 'adding'
slc_ofs += slc_n
dbi.Session.commit()
print fail_count, sxs_count
def fill_db( name = '16s', reset = True):
dbi = cbdb.getName(name, tables = get_tables(),
reset = np.mod(reset,2))
count = 0
for p in paths:
fopen = open(p)
a = dbi.Alignment()
dbi.Session.add(a)
dbi.Session.commit()
for rec in SeqIO.parse(fopen, 'genbank'):
f0 = rec.features[0]
if f0.type == 'source':
source_taxon = f0.qualifiers['db_xref'][0][6:]
source_organism=f0.qualifiers['organism'][0]
else:
source_taxon = None
source_organism = None
seq = dbi.Sequence(name = rec.name,
file_name = p,
file_offset = fopen.tell(),
sequence = rec.seq.__str__(),
source_taxon = source_taxon,
source_organism = source_organism,
gb_accession = rec.id,
annotations = rec.annotations.__str__(),
alignmentid = 0)
dbi.Session.add(seq)
if np.mod(count, 1000) == 0:
print count, p , seq.source_organism
dbi.Session.commit()
count += 1
dbi.Session.commit()
def makeRank(self, rank = 'phylum', subtree = None):
#Get the subtree and db connections to build meta for
tree = subtree if subtree != None else self.t
dbi = cbdb.getName('taxdmp')
print 'Fetching taxonomic nodes from the db'
#Get the terminal nodes and corresponding ncbi taxa
terms = [t for t in tree.get_terminals() if t.m.has_key('taxid')]
nodes = [dbi.S.q(dbi.Node).filter_by(id = t.m['taxid']).scalar()
for t in terms]
#endpoints for parental iteratiion
taxa = ncbi.get_rank(rank)
root = ncbi.get_root()
print 'Computing terminal node mappings for taxon: {0}'.format(rank)
bar = pbar.simple(len(nodes)); bar.start()
node_taxa = list(nodes)
get_p_iter = lambda: \
node_taxa[idx] == None and True \
or node_taxa[idx] in taxa and True \
or node_taxa[idx] == root and True \
or node_taxa.__setitem__(idx,node_taxa[idx].parent) \
or node_taxa[idx]
for idx, v in enumerate(node_taxa):
bar.update(idx);
par = list(iter(get_p_iter, True))[-1] if v else None
terms[idx].m[rank] = par.id if par in taxa else None
bar.finish()
print 'Done!'
def getGenealogy(node):
root_node =taxRoot()
dbi = cbdb.getName('taxdmp')
path = []
cur = node
while cur != root_node:
path.append(cur)
cur = cur.parent
path.append(cur)
return path[::-1]
def fill_all_rdb16s(reset = True):
paths = []
for r, ds, fs in os.walk(config.dataPath('alignments/16s')):
for f in fs:
if '.gbk' in f:
paths.append(os.path.join(r,f))
cbdb = compbio.projects.cbdb
dbi = cbdb.getName('16s',
tables = get_tables(),
reset = np.mod(reset, 2))
last_ofs = 0
for p in paths:
fopen = open(p)
a = dbi.Alignment(file_name =config.dataURL(p))
dbi.Session.add(a)
dbi.Session.commit()
count = 0
for rec in SeqIO.parse(fopen, 'genbank'):
try:
src_taxon = rec.features[0].qualifiers['db_xref'][0][6:]
except Exception, e:
src_taxon = None
ann = sjson.dumps(rec.annotations, default = lambda x: x.__str__())
seq = dbi.Sequence(name = rec.name,
file_name = p,
file_offset = last_ofs,
sequence = rec.seq.__str__(),
gb_accession = rec.id,
gb_accession_version = 1,
gb_id = None,
annotations = ann,
alignment = a,
source_taxon = src_taxon
)
dbi.Session.add(seq)
last_ofs = fopen.tell()
if np.mod(count, 1000) == 0:
print count, p, seq.source_organism
dbi.Session.commit()
count += 1
dbi.Session.commit()
def fill_db( reset = True):
dbi = cbdb.getName('taxdmp', tables = get_tables(), reset = np.mod(reset, 2))
filepath = config.dataPath('ncbi/taxdmp')
maps = get_maps()
record_sep = '\t|\n'
col_sep = '\t|\t'
colfun = lambda x: unicode(x, errors = 'replace').replace(record_sep, '').split(col_sep)
record_iterfun = lambda x: x.xreadlines()
fill_tables = {'Gencode':'gencode.dmp',
'Node':'nodes.dmp',
'Name':'names.dmp',
'Citation':'citations.dmp'}
count = 0
for k,v in fill_tables.iteritems():
fopen = open(os.path.join(filepath, v))
fsize = os.path.getsize(os.path.join(filepath,v))
mapped_class = dbi.__dict__[k]
mapped_columns = maps[k]
l0 = ''
for l in record_iterfun(fopen):
count += 1
l0+=l
if l0[-3:] == record_sep :
l = l0
l0 = ''
else: continue
cols = colfun(l)
cls = mapped_class(**dict(map(lambda (x,y): (x,cols[y]),
mapped_columns.iteritems())))
dbi.Session.merge(cls)
if np.mod(count, 1000) == 0:
dbi.Session.commit()
print k, v, count, cols, '{0:4}%'.format(100 * float(fopen.tell()) / fsize)
dbi.Session.commit()
return
def fill_from_rfam_stk( p, reset = True):
cbdb = compbio.projects.cbdb
aname = os.path.basename(p)
dbi = cbdb.getName(aname, tables = get_tables(),
reset = np.mod(reset,2))
fopen = open(p)
a = dbi.Alignment(file_name = aname)
dbi.Session.add(a)
dbi.Session.commit()
count = 0
for rec in SeqIO.parse(fopen, 'stockholm'):
acc = rec.annotations['accession']
accidv, accrange = acc.split('/')
acv_split = accidv.split('.')
accid = acv_split[0]
accid_version = (lambda x: len(x) == 1 and 1 or x[1])(acv_split)
ann = sjson.dumps(rec.annotations, default = lambda x: x.__str__())
seq = dbi.Sequence(name = rec.name,
file_name = p,
file_offset = fopen.tell(),
sequence = rec.seq.__str__(),
gb_accession = accid,
gb_accession_version = accid_version,
gb_accession_range = accrange,
gb_id = None,
annotations = ann,
alignment = a
)
dbi.Session.add(seq)
if np.mod(count, 100) == 0:
print count, p , seq.source_organism
dbi.Session.commit()
count += 1
dbi.Session.commit()
def investigatePhylum(self,
aliname = 'group2.stk',
p_node = None, **kwargs):
if not p_node: p_node = ncbi.taxon_with_name('phylum', 'Thermotogae')
ali_seqs = ali.get_seqs(aliname, **mem.sr(kwargs))
ali_nodes = array(ali.get_taxnodes(aliname, **mem.sr(kwargs)))
ali_phyla = array(ali.get_taxon_forall(aliname,**mem.sr(kwargs, rank = 'phylum')))
ali_inds = nonzero(equal(ali_phyla, p_node))[0]
leaf_terminals = self.t.get_terminals()
leaf_nodes = array(self.leafNodes(**mem.sr(kwargs)))
leaf_phyla = array(self.getTaxon('phylum', **mem.sr(kwargs)))
leaf_inds = nonzero(equal(leaf_phyla, p_node))[0]
ap_sub = ali_phyla[ali_inds]
lp_sub = leaf_phyla[leaf_inds]
ag_sub = array(ali.get_taxon_forsome(ali_nodes[ali_inds],'genus','thermo',
**mem.sr(kwargs)))
lg_sub = array(self.getTaxon('genus', **mem.sr(kwargs)))[leaf_inds]
as_sub = array(ali.get_taxon_forsome(ali_nodes[ali_inds], 'species', 'thermo'))
ls_sub = array(self.getTaxon('species',**mem.sr(kwargs)))[leaf_inds]
db16 = cbdb.getName('16s')
a_16s= [ db16.S.q(db16.Sequence).
filter_by(source_taxon = n.id).all() for n in ali_nodes[ali_inds]]
l_16s= [ db16.S.q(db16.Sequence).
filter_by(source_taxon = n.id).all() for n in leaf_nodes[leaf_inds]]
#fill any empty nodes... (those lacking 16s rRNA)
for idx, elt in enumerate(a_16s):
cur_node= ali_nodes[ali_inds[idx]]
while not elt:
cur_node = cur_node.parent
elt.extend(db16.S.q(db16.Sequence).filter_by(source_taxon = cur_node.id).all())
for idx, elt in enumerate(l_16s):
cur_node= leaf_nodes[leaf_inds[idx]]
while not elt:
cur_node = cur_node.parent
elt.extend(db16.S.q(db16.Sequence).filter_by(source_taxon = cur_node.id).all())
all_lens = dict([ (k, [len(list( e)) for e in seqlist] )
for seqlist,k in [[a_16s,'a_16s'],[l_16s,'l_16s']]])
leaf_sns = [ SeqNode(lg_sub[i],ls_sub[i] , leaf_nodes[idx],
[(x.sequence,x.gb_id, x.source_taxon, ncbi.get_node(x.source_taxon).rank)
for x in l_16s[i]],
src = leaf_terminals[idx],
node_id = 'btol:default:{0}'.format(leaf_terminals[idx].m['id']))
for i, idx in enumerate(leaf_inds)]
ali_sns = [ SeqNode(ag_sub[i],as_sub[i] , ali_nodes[idx],
[( x.sequence,x.gb_id , x.source_taxon, ncbi.get_node(x.source_taxon).rank)
for x in a_16s[i]],
src = ali_seqs[idx],
node_id = 'ali:{0}:{1}'.format(aliname,ali_seqs[idx].id))
for i, idx in enumerate(ali_inds)]
return list(it.chain(leaf_sns, ali_sns))
def setLeafNodes(**kwargs):
all_leaves = self.t.get_terminals()
dbi = cbdb.getName('taxdmp')
all_nodes = [ ncbi.get_node(l.m['taxid'],dbi)
if 'taxid' in l.m.keys() else None for l in all_leaves]
return all_nodes
def taxRoot():
tax_dbi = cbdb.getName('taxdmp')
root_node = tax_dbi.Session.query(tax_dbi.Name).filter_by(name_txt = 'root').one().node
return root_node
def rna4gbid( gbid, dbname = '16s'): #print 'giving a random RNA because the taxonomy atabase is not yet created!' dbi = cbdb.getName(dbname) seq_num= floor(1000 * random.random()) seq = dbi.Session.query(dbi.Sequence)[seq_num].sequence return seq
def get_root():
dbi = cbdb.getName('taxdmp')
return dbi.S.q(dbi.Node).\
filter_by(id = 1).one()
def get_rank(rankname = 'phylum'):
dbi = cbdb.getName('taxdmp')
rank = dbi.S.q(dbi.Node).\
filter_by(rank = rankname).all()
return rank
def get_node(nodeid, dbi = None):
if dbi == None: dbi = cbdb.getName('taxdmp')
return dbi.S.q(dbi.Node).filter_by(id=nodeid).scalar()
def set_seqs(**kwargs): cbdb = compbio.projects.cbdb dbname = kwargs['dbname'] dbi = cbdb.getName(dbname) nodes = dbi.S.q(dbi.Sequence).all() return nodes
def fill_db( name = 'bacterial_genomes', reset = False,
postgres = False, host = 'broad'):
dbi = cbdb.getName(
name,
postgres = postgres,
tables = get_tables(),
reset = np.mod(reset, 2),
host = host)
paths = []
for r,ds, fs in os.walk('/Volumes/ganymede/all.gbk/'):
for f in fs:
if 'gbk' in f: paths.append(os.path.join(r, f))
count = 0
for p in paths:
if count < 1668:
count += 1
continue
count += 1
fopen = open(p)
for rec in SeqIO.parse(fopen, 'genbank'):
f0 = rec.features[0]
if f0.type == 'source':
source_taxon = f0.qualifiers['db_xref'][0][6:]
source_organism=f0.qualifiers['organism'][0]
else:
source_taxon = None
source_organism = None
fa_seqpath = 'genomes/'+rec.id+'.fa'
fa_sequrl = config.dataURL(fa_seqpath)
fa_seqfile = config.dataPath(fa_sequrl)
fopen = open(fa_seqfile,'w')
SeqIO.write(rec,fopen, 'fasta')
fopen.close()
adds = []
genome = dbi.Genome(name = rec.name,
seq_url =fa_sequrl,
source_taxon = source_taxon,
source_organism = source_organism,
gb_accession = rec.id,
annotations = rec.annotations.__str__())
#adds.append(genome)
print 'adding genome ' + source_organism
dbi.Session.add(genome)
print 'commiting update '
dbi.Session.commit()
print 'genome added! '
for f in rec.features:
feature = dbi.Feature(type = f.type,
start = f.location.start.position,
start_ext = f.location.start.extension,
end = f.location.end.position,
end_ext = f.location.end.extension,
strand = f.strand,
genomeobj = genome)
#print 'adding feature ' + f.type
#dbi.Session.add(feature)
adds.append(feature)
for k,v in f.qualifiers.iteritems():
q = dbi.Qualifier(key = k,
value = v.__str__(),
featureobj = feature)
#dbi.Session.add(q)
adds.append(q)
for sf in f.sub_features:
sub = dbi.SubFeature(type = sf.type,
start = sf.location.start.position,
start_ext = sf.location.start.extension,
end =sf.location.end.position,
end_ext = sf.location.end.extension,
strand = sf.strand,
featureobj = feature)
adds.append(sub)
#dbi.Session.add(sub)
for k,v in sf.qualifiers.iteritems():
q = dbi.Qualifier(key = k,
value = v.__str__(),
subfeatureobj = sf)
#Session.add(q)
adds.append(q)
dbi.Session.add_all(adds)
if np.mod(count, 2) == 0:
print count
#print count, p , seq.source_organism
print 'committing update'
dbi.Session.commit()
print 'update commited!'
dbi.Session.commit()
