def definesets(opt):
assert opt.bfile is not None, 'Please specify a bfile.'
""" precomputing the windows """
if opt.wfile is None:
opt.wfile = os.path.split(opt.bfile)[-1] + '.%d' % opt.window_size
warnings.warn('wfile not specifed, set to %s' % opt.wfile)
t0 = time.time()
(bim, fam, G) = read_plink(opt.bfile, verbose=False)
if opt.sliding_wind:
print('Precomputing windows')
sets = sets_from_bim(bim,
size=opt.window_size,
step=opt.step,
chrom=opt.chrom,
minSnps=opt.minSnps,
maxSnps=opt.maxSnps)
elif opt.filter_sets:
assert opt.iwfile is not None, 'Please specify a iwfile.'
print('Filtering windows')
pdb.set_trace()
sets = annotate_sets(sets0,
bim,
minSnps=opt.minSnps,
maxSnps=opt.maxSnps)
sets.to_csv(opt.wfile + '.wnd', sep='\t', index=False)
print(('Number of variants:', G.shape[0]))
print(('Number of windows:', sets.shape[0]))
t1 = time.time()
print(('.. finished in %s seconds' % (t1 - t0)))
def _load(self):
(bim, fam, bed) = read_plink(self._prefix, verbose=False)
self._snpinfo = bim
self._ind_info = fam
self._geno = bed
# convert to pandas dataframe and export
out = pd.DataFrame(out,
columns=['setid', 'chrom', 'start', 'end', 'nsnps'])
return out
def calc_nsnps(pos, start, end):
nsnps = np.zeros(start.shape[0], dtype=int)
for i, r in enumerate(zip(start, end)):
nsnps[i] = ((pos >= r[0]) & (pos < r[1])).sum()
return nsnps
if __name__ == '__main__':
import os
bedpath = 'data/chrom22_subsample20_maf0.10'
if not os.path.exists(bedpath + '.bim'):
os.system('wget http://www.ebi.ac.uk/~casale/data.zip')
os.system('unzip data.zip')
(bim, fam, G) = read_plink(bedpath)
# sets_from_bim
sets = sets_from_bim(bim)
# annotate_sets
sets1 = annotate_sets(sets, bim, minSnps=100, maxSnps=200)