from cruzdb import Genome
from cruzdb.sequence import sequence
# mirror the neede tables from UCSC to a local sqlite db
local = Genome('hg19').mirror(('refGene', 'targetScanS'), 'sqlite:///hg19.mirna.db')
# connect to the newly created local sqlite database instance.
refseq_ids = []
# iterate over the coding in refGene
for gene in (rgene for rgene in local.refGene if rgene.is_coding):
if None in gene.utr3: continue # skip genes with no UTR
utr_start, utr_end = gene.utr3
# query the targetScan miRNA table with efficient bin query
sites = local.bin_query('targetScanS', gene.chrom, utr_start, utr_end)
# print BED file of genes whose 3'UTR contains a miR-96 target site
# with a score > 85.
if any("miR-96" in s.name and s.score > 85 for s in sites):
refseq_ids.append(gene.name) # save the refSeq for later GO analysis
# gene is a python object but its string representation is BED format
# we also print out the UTR sequence.
print gene, sequence('hg19', gene.chrom, utr_start, utr_end)
# open a webbrowser to show enrichment of the genes we've selected in DAVID
Genome.david_go(refseq_ids)
print g.upstream("refGene", last, k=6)
1 / 0
seed(1)
istart = 12345
iend = 386539
qall = list(g.refGene.all())
#while True:
for iend in (randrange(istart, 65555555) for i in range(100)):
t = time.time()
q = g.bin_query('refGene', 'chr1', istart, iend)
a = list(q)
print len(a)
print time.time() - t
#"""
t = time.time()
refGene = g.refGene
rg = refGene.table()
q = g.session.query(rg).filter(rg.c.chrom == "chr1", rg.c.txStart <= iend,
rg.c.txEnd >= istart)
q = refGene.filter(rg.c.chrom == "chr1", rg.c.txStart <= iend,
rg.c.txEnd >= istart)
b = list(q)
print len(b)
if not op.exists(fname):
fhout = open(fname, 'w')
hg18.annotate(lamina(), ('refGene', ), feature_strand=True, in_memory=True, parallel=True, out=fhout)
fhout.close()
for cutoff in (0.90, 0.95):
fh = open('/tmp/genes-%.2f.txt' % cutoff, 'w')
for d in reader(fname):
if float(d['value']) < cutoff: continue
if d['refGene_distance'] == '0' or \
d['refGene_distance'].startswith("0;"):
print >>fh, "\n".join(d['refGene_name'].split(";"))
fh.close()
cutoff = 0.90
fh = open('/tmp/genes-overlap-complete.txt', 'w')
for d in (l for l in reader(lamina()) if float(l['value']) > cutoff):
if float(d['value']) < cutoff: continue
start, end = map(int, (d['start'], d['end']))
res = hg18.bin_query('refGene', d['chrom'], start, end).all()
if len(res) == 0: continue
for r in res:
# genes completely contained within an LAD
if start <= r.start and end >= r.end:
print >>fh, r.gene_name
1/0
seed(1)
istart = 12345
iend = 386539
qall = list(g.refGene.all())
#while True:
for iend in (randrange(istart, 65555555) for i in range(100)):
t = time.time()
q = g.bin_query('refGene', 'chr1', istart, iend)
a = list(q)
print len(a)
print time.time() - t
#"""
t = time.time()
refGene = g.refGene
rg = refGene.table()
q = g.session.query(rg).filter(rg.c.chrom == "chr1", rg.c.txStart
<= iend, rg.c.txEnd >= istart)
q = refGene.filter(rg.c.chrom == "chr1", rg.c.txStart
<= iend, rg.c.txEnd >= istart)
b = list(q)
# MySQLdb stuff:
# sudo apt-get install mysql-server
# sudo apt-get install libmysqlclient-dev # gives us mysql_config
# FINALLY download MySQLdb source, do process described in INSTALL file
hg19 = Genome('hg19')
INPUTFILE = "suggestive.pheno_simple.covar_none.test_wald.csv"
filereader = csv.reader(open(INPUTFILE))
chrom_i = None
pos_i = None
for i, line in enumerate(filereader):
if i == 0:
# CHROM POS REF ALT N_INFORMATIVE Test Beta SE Pvalue PVALUE
chrom_i = line.index('CHROM')
pos_i = line.index('POS')
continue
chrom = 'chr' + str(line[chrom_i])
pos = int(line[pos_i])
start = pos - 50 # kind of arbitrary search 50 back 50 forward.
end = pos + 50
genes = hg19.bin_query('refGene', chrom, start, end)
## formatting the output
basic_str = ' '.join(map(str, [chrom, start, end]))
padding = 30 - len(basic_str)
if padding < 0:
padding = 1
gene_string = ' '.join(set(g.name2 for g in genes))
print basic_str + ' ' * padding + gene_string
