def combine_corrs(e1, e2, allprots, combine_func, default_val=None):
# we combine the symmetric correlation matrices using the specified
# element-wise function. function examples: max, sum
# we use the specified ordering of elements in allprots
default_val = default_val if default_val else -1 if combine_func.__name__.find("max") > -1 else 0
nprots = len(allprots)
corr = np.matrix(np.zeros((nprots, nprots)))
dprots1 = ut.list_inv_to_dict(e1.prots)
dprots2 = ut.list_inv_to_dict(e2.prots)
for row, p1 in enumerate(allprots):
for col, p2 in enumerate(allprots):
val1 = e1.corr[dprots1[p1], dprots1[p2]] if p1 in dprots1 and p2 in dprots1 else default_val
val2 = e1.corr[dprots2[p1], dprots2[p2]] if p1 in dprots2 and p2 in dprots2 else default_val
corr[row, col] = combine_func(val1, val2)
return corr
def prot_counts_pep2prots(peplist, only_uniques, pep2prots):
"""
- pep2prots: use supplied dict of {peptide: set(proteinids)} instead of the
protein ids on the lines in the pep_list file, in which case sum up the
counts for a peptide-spectral combination.
"""
assert only_uniques, "Only handles only_uniques=True so far."
exclude_peps = (set([pep for pep,prots in pep2prots.items() if len(prots)>1])
if only_uniques else set([]))
print "%s non-unique peptides to exclude." % len(exclude_peps)
pep_samp_counts = defaultdict(float)
for _,sample,pep,count in peplist:
if pep not in exclude_peps:
pep_samp_counts[(pep,sample)] += float(count)
# Currently ignoring peptides without a mapping.
prots = sorted(list(reduce(set.union, [pep2prots[pep] for (pep,_),_ in
pep_samp_counts.items() if pep in pep2prots])))
samples = sorted(list(set(ut.i1(peplist))))
print "%s unique proteins. %s samples." % (len(prots), len(samples))
dprots, dsamples = [ut.list_inv_to_dict(lst) for lst in prots, samples]
counts = np.zeros((len(prots), len(samples)), dtype='float32')
for (pep,sample),count in pep_samp_counts.items():
if pep in pep2prots: # Currently ignoring peptides without a mapping.
assert len(pep2prots[pep])==1, "Non-unique peptide found"
counts[dprots[list(pep2prots[pep])[0]], dsamples[sample]] += count
totals = counts.sum(axis=1)
nonzero = totals > 0
prots, totals = [list(np.array(lst)[nonzero]) for lst in prots, totals]
counts = counts[nonzero,:]
return prots, samples, counts, totals
def __init__(self, filename, sp_base="Hs", norm_rows=False, norm_cols=False):
e = load_elution(filename)
self.prots = e.prots
self.filename = e.filename
self.normarr = ut.normalize_fracs(e.mat, norm_rows=norm_rows, norm_cols=norm_cols)
self.pinv = ut.list_inv_to_dict(e.prots)
sp_target = ut.shortname(e.filename)[:2]
self.baseid2inds = sc.orth_indices(sp_base, sp_target, e.prots, False)
def __init__(self, filename, sp_base='Hs', norm_rows=False, norm_cols=False):
e = load_elution(filename)
self.prots = e.prots
self.filename = e.filename
self.normarr = ut.normalize_fracs(e.mat, norm_rows=norm_rows,
norm_cols=norm_cols)
self.pinv = ut.list_inv_to_dict(e.prots)
sp_target = ut.shortname(e.filename)[:2]
self.baseid2inds = sc.orth_indices(sp_base, sp_target, e.prots, False)
def combine_corrs(e1, e2, allprots, combine_func, default_val=None):
# we combine the symmetric correlation matrices using the specified
# element-wise function. function examples: max, sum
# we use the specified ordering of elements in allprots
default_val = default_val if default_val else -1 if \
combine_func.__name__.find('max') > -1 else 0
nprots = len(allprots)
corr = np.matrix(np.zeros((nprots,nprots)))
dprots1 = ut.list_inv_to_dict(e1.prots)
dprots2 = ut.list_inv_to_dict(e2.prots)
for row,p1 in enumerate(allprots):
for col,p2 in enumerate(allprots):
val1 = e1.corr[dprots1[p1], dprots1[p2]] if p1 in dprots1 and p2 in \
dprots1 else default_val
val2 = e1.corr[dprots2[p1], dprots2[p2]] if p1 in dprots2 and p2 in \
dprots2 else default_val
corr[row,col] = combine_func(val1, val2)
return corr
def ppis_add_sp_ppis(ppis, arrfeats):
idict = ut.list_inv_to_dict(((r[0],r[1]) for r in arrfeats))
newppis = []
cols = [n for n in arrfeats.dtype.names if n[2:].startswith('_ppi_score')]
for p in ppis:
index = idict[(p[0],p[1])]
sp_ppis = arrfeats[index:index+1][cols][0]
newppis.append(tuple(p) + tuple(sp_ppis))
return newppis, cols
def elut_gene_maxes(elutfs, geneids):
d = {}
for f in elutfs:
e = el.load_elution(f)
prots_inv = ut.list_inv_to_dict(e.prots)
for gid in geneids:
if gid in prots_inv:
d.setdefault(f,{})[gid] = np.max(e.mat[prots_inv[gid]])
return d
def ppis_add_splist(ppis, arrfeats, cutoff):
idict = ut.list_inv_to_dict(((r[0],r[1]) for r in arrfeats))
newppis = []
print "Cutoff: %s" % cutoff
for p in ppis:
index = idict[(p[0],p[1])]
splist = fe.passing_species_separate(arrfeats[index:index+1], cutoff)
newppis.append(tuple(p) + tuple(splist))
sps = fe.species_list(arrfeats.dtype.names[3:])
return newppis, sps
def filter_matching_elution(edata, efilter, remove_data_ending=".map"):
"""
Use efilter as a mask on edata, setting edata to 0 based on efilter being 0
"""
newmat = np.matrix(np.zeros(edata.mat.shape))
# First create the column-matched array from filter to data
inv_fracs = ut.list_inv_to_dict(efilter.fractions)
data_fracs = [f.replace(remove_data_ending, "") for f in edata.fractions]
arrfilt = np.zeros((efilter.mat.shape[0], edata.mat.shape[1]))
for i, f in enumerate(data_fracs):
arrfilt[:, i] = np.asarray(efilter.mat)[:, inv_fracs[f]] if f in inv_fracs else np.zeros(efilter.mat.shape[0])
# Then go row-by-row
filter_map = ut.list_inv_to_dict(efilter.prots)
for i, g in enumerate(edata.prots):
if g in filter_map:
newmat[i, :] = np.asarray(edata.mat)[i, :] * (arrfilt[filter_map[g], :] > 0).astype(int)
else:
newmat[i, :] = np.zeros(edata.mat.shape[1])
return newmat
def complex_arr(cxs, prots):
arr = np.zeros((len(prots),len(prots)))
ints_dict = co.corum_ints_duped([(i,ps) for i,ps in enumerate(cxs)])
p_inds = ut.list_inv_to_dict(prots)
for p,partners in ints_dict.items():
if p in p_inds:
for partner in partners:
if partner in p_inds:
arr[p_inds[p], p_inds[partner]] = 1
return arr
def profiles_cxs(e, cxs, **kwargs):
# blue/yellow/red map: 'jet'
defaults = {'interpolation': 'nearest', 'cmap':'hot', 'vmin':1}
kwargs = ut.dict_set_defaults(kwargs, defaults)
arr = np.array(e.mat)
dinds = ut.list_inv_to_dict(e.prots)
useps = [p for c in cxs for p in c]
useinds = [dinds[p] for p in useps if p in dinds]
vals = np.clip(np.log2(arr[useinds,:]),0,100)
imshow(vals, **kwargs)
return vals
def filter_matching_elution(edata, efilter, remove_data_ending='.map'):
"""
Use efilter as a mask on edata, setting edata to 0 based on efilter being 0
"""
newmat = np.matrix(np.zeros(edata.mat.shape))
# First create the column-matched array from filter to data
inv_fracs = ut.list_inv_to_dict(efilter.fractions)
data_fracs = [f.replace(remove_data_ending,"") for f in edata.fractions]
arrfilt = np.zeros((efilter.mat.shape[0], edata.mat.shape[1]))
for i,f in enumerate(data_fracs):
arrfilt[:,i] = (np.asarray(efilter.mat)[:,inv_fracs[f]] if f in
inv_fracs else np.zeros(efilter.mat.shape[0]))
# Then go row-by-row
filter_map = ut.list_inv_to_dict(efilter.prots)
for i,g in enumerate(edata.prots):
if g in filter_map:
newmat[i,:] = (np.asarray(edata.mat)[i,:] *
(arrfilt[filter_map[g],:] > 0).astype(int))
else:
newmat[i,:] = np.zeros(edata.mat.shape[1])
return newmat
def orth_indices(sp_base, sp_target, prot_list, remove_multi_base):
"""
Using appropriate orthology, take a list of target species gene ids
(corresponding to rows in the target species score matrix), and
return a dict mapping base species gene ids to (sets of) indices in that
list and therefore to (sets of) row/column indices in the square
interaction score matrix.
"""
targ2inds = dict([(k,set([v]))
for k,v in ut.list_inv_to_dict(prot_list).items()])
if sp_base == sp_target:
return targ2inds
else:
base2targ = orth.odict(sp_base, sp_target)
if remove_multi_base:
base2targ = remove_multi_keys(base2targ)
base2inds = ut.compose_dict_sets(base2targ, targ2inds)
base2inds = dict([(k,v) for k,v in base2inds.items() if len(v)>0])
return base2inds
def orth_indices(sp_base, sp_target, prot_list, remove_multi_base):
"""
Using appropriate orthology, take a list of target species gene ids
(corresponding to rows in the target species score matrix), and
return a dict mapping base species gene ids to (sets of) indices in that
list and therefore to (sets of) row/column indices in the square
interaction score matrix.
"""
targ2inds = dict([(k,set([v]))
for k,v in ut.list_inv_to_dict(prot_list).items()])
if sp_base == sp_target:
return targ2inds
else:
base2targ = orth.odict(sp_base, sp_target)
if remove_multi_base:
base2targ = remove_multi_keys(base2targ)
base2inds = ut.compose_dict_sets(base2targ, targ2inds)
base2inds = dict([(k,v) for k,v in base2inds.items() if len(v)>0])
return base2inds
def prot_counts(peplist, only_uniques):
"""
Deprecated 20130628. pep_list doesn't seem consistent for protein
assignments.
mainly returns the counts array with protein quantations.
- exclude_peps: set of peptides to exclude, probably from non_unique_peps.
"""
print "**Deprecated 20130628**"
exclude_peps = non_unique_peps(peplist) if only_uniques else set([])
prots,samples = [sorted(list(set(lst)))
for lst in zip(*[i[:2] for i in peplist])]
dprots, dsamples = [ut.list_inv_to_dict(lst) for lst in prots, samples]
counts = np.zeros((len(prots), len(samples)), dtype='float32')
for prot_peplist, sample, pep, count in peplist:
if pep not in exclude_peps:
counts[dprots[prot],dsamples[sample]] += float(count)
totals = counts.sum(axis=1)
nonzero = totals > 0
prots, totals = [list(np.array(lst)[nonzero]) for lst in prots, totals]
counts = counts[nonzero,:]
return prots, samples, counts, totals
def pair2ind(items):
return ut.list_inv_to_dict(((x[0],x[1]) for x in items))
