def get_expression(**kwargs):
atype = "wormtile"
simple = wp.get_simple_thr(**mem.rc(kwargs, atype=atype, dthr=3000, dsign=-1, sthr=1e-5))
idfun = lambda g: g.qualifiers["db_xref"][1][9:]
genes = dict([(k, set([g for g in v["gnames"]])) for k, v in simple.iteritems()])
gene_union = set([])
for glist in genes.values():
gene_union.update(glist)
crofs = wp.chromosome_offsets()
gene_info = wp.gene_info(**mem.rc(kwargs))
genome_coords = dict([(k, gene_info[k]["genomestart"]) for i, k in enumerate(gene_union)])
gene_idxs = dict([(k, i) for i, k in enumerate(gene_union)])
gnames = list(gene_union)
gene_srtlist = argsort([e[1] for e in sorted(gene_idxs.iteritems(), key=lambda x: genome_coords[x[0]])])
gene_srtidxs = dict([(k, gene_srtlist[e]) for k, e in gene_idxs.iteritems()])
na = len(simple)
ng = len(gene_union)
gc = zeros((na, ng))
assay_coords = dict([(k, i) for i, k in enumerate(simple.keys())])
gene_counts = dict(
[(k, dict([(k2, len(list(g2))) for k2, g2 in it.groupby(sorted(v["gnames"]))])) for k, v in simple.iteritems()]
)
for k, v in gene_counts.iteritems():
for k2, v2 in v.iteritems():
gc[assay_coords[k], gene_srtidxs[k2]] = v2
return 1 * greater(gc, 0)
def set_easy0(**kwargs):
atype = kwargs.get("atype")
simple = wp.get_simple_thr(atype=atype, dthr=1500, dsign=-1, sthr=1e-4)
raise Exception()
score_soft_cut = -136
score_hard_cut = -90
sids = wu.symbol_ids()
prop_tuples = []
for tf, props in simple.iteritems():
# for now, remove tfs that are not mappable
if not tf in sids.keys():
continue
ssrt = argsort(props["scores"])
lscores = log10(props["scores"][ssrt])
easy = nonzero(less(lscores, score_soft_cut))[0]
medium = nonzero(greater(lscores, score_soft_cut) * less(lscores, score_hard_cut))[0]
# generous_edges = concatenate([easy,medium])
prop_tuples.append(
[
(tf, props["genes"][g], -(score_hard_cut - lscores[g]) / (score_soft_cut - score_hard_cut))
for g in medium
]
)
prop_tuples.append([(tf, props["genes"][g], 1) for g in easy])
edgelist = array(list(it.chain(*prop_tuples)))
edges = [(sids[e[0]], e[1].qualifiers["db_xref"][1][9:], e[2]) for e in edgelist]
return edges
def peak_thr_histograms(**kwargs):
'''histograms of score and distance'''
dthr = 1500
sthr = 1e-2
dsign = -1
simple = wp.get_simple_thr(**mem.rc(kwargs,
dthr = dthr,
sthr = sthr,
dsign = dsign
)
)
min_score = -1
max_score = -1
for k,v in simple.iteritems():
smax = np.max(v['scores'])
if max_score == -1 or smax > max_score:
max_score = smax
smin = np.min(v['scores'])
if min_score == -1 or smin < min_score:
min_score = smin
lrange = [int(floor(log10(min_score))), int(ceil(log10(max_score)))]
sbin_mids = range(lrange[0],lrange[1]+1)
nsb = len(sbin_mids)
sbins = zeros((nsb))
dbin_size = 50
dbin_mids = range(-dthr, dthr, dbin_size)
ndb = len(dbin_mids)
dbins = zeros(( ndb))
for k,v in simple.iteritems():
for d in v['dists']:
dbins[int(d + dthr)/dbin_size] += 1
for s in v['scores']:
sbins[int(log10(s) - lrange[0])] += 1
f= myplots.fignum(1,(10,6))
ax = f.add_subplot(121)
ax.set_ylabel('log 10 counts')
ax.set_xlabel('distance')
ax.set_title('simplified tss distances (d<{0})'.format(dthr))
ax.plot(dbin_mids,log10(dbins), color = 'black')
f.savefig(myplots.figpath('chip_simple_distance.pdf'))
ax = f.add_subplot(122)
ax.set_title('simplified tss scores (s<{0:2.2})'.format(sthr))
ax.set_ylabel('log10 counts')
ax.set_xlabel('log10 peak score')
ax.plot(sbin_mids,log10(sbins), color = 'black')
f.savefig(myplots.figpath('chip_simple_scores.pdf'))
