def peakcros(args, xargs):
with open(args.peakfile) as fin:
peaks = map(lambda t: t.split(), fin)
peaks = peakset(map(lambda t: [float(t[0]), float(t[1]), t[2]], peaks))
lrbt = [args.bleft, args.bright, args.bbottom, args.btop]
for sample in eval(args.holonums):
holo = hsqc(os.path.join(args.folder, str(sample) + ".nv"))
holopeaks = peakset(
holo.xyzconv(holo.peak_search(args.num_peaks, lrbt)))
print "Sample #" + str(sample).ljust(
3) + ": %d peaks matched" % holopeaks.matchcount(
peaks, args.dcutoff, args.scutoff)
def peakmatch(args, xargs):
with open(args.peakfile) as fin:
peaks = map(lambda t: t.split(), fin)
Np = len(peaks)
peaks = peakset(map(lambda t: [float(t[0]), float(t[1]), t[2]], peaks))
lrbt = [args.bleft, args.bright, args.bbottom, args.btop]
allpms, hps = [], []
for sample in eval(args.holonums):
holo = hsqc(os.path.join(args.folder, str(sample) + ".nv"))
holopeaks = peakset(
holo.xyzconv(holo.peak_search(args.num_peaks, lrbt)))
samplenum = str(sample).ljust(3)
print "Sample #" + samplenum + " processed"
pms = holopeaks.peakmatch(peaks)
allpms.append(list(pms[1]))
hps.append(holopeaks)
x = array(allpms).flatten()
x.sort()
N = len(x)
n = arange(1, N).astype(float)
sigmaD = sqrt(0.5 * x[:N / 2]**2 / log(N / (N - n[:N / 2]))).mean()
print "Estimated peak position noise level %.3f" % (sigmaD)
zcut = bonferroni(args.zcutoff, Np)
print "Requested Z-score cutoff (%.2f) adjusted to %.2f (Bonferroni correction)" % (
args.zcutoff, zcut)
dcut = zcut * sigmaD
print "Peak mismatch cutoff set to %.3f ppm" % dcut
for (i, sample) in enumerate(eval(args.holonums)):
holopeaks = hps[i]
pms = holopeaks.peakmatch(peaks)
samplenum = str(sample).ljust(3)
print "Sample #" + samplenum + ": %d peaks shifted" % (
Np - holopeaks.matchcount(peaks, dcut, args.scutoff))
if args.print_peakmatch:
pms = sorted(zip(*(pms + [list(zip(*peaks.peaks)[2])])),
key=lambda x: x[1],
reverse=True)
print '\n'.join(
map(lambda x: samplenum + "%7.3f %6.2f %s" % x[1:],
filter(lambda xx: xx[1] >= dcut, pms)))
def peakmxy(args, xargs):
with open(args.peakfile) as fin:
peaks = map(lambda t: t.split(), fin)
Np = len(peaks)
peaks = peakset(map(lambda t: [float(t[0]), float(t[1]), t[2]], peaks))
lrbt = [args.bleft, args.bright, args.bbottom, args.btop]
allxy, hps = [], []
for sample in eval(args.holonums):
holo = hsqc(os.path.join(args.folder, str(sample) + ".nv"))
holopeaks = peakset(
holo.xyzconv(holo.peak_search(args.num_peaks, lrbt)))
hps.append(holopeaks)
allxy.append(holopeaks.peakmxy(peaks))
samplenum = str(sample).ljust(3)
sys.stderr.write("Sample #" + samplenum + " processed\n")
xx, yy = array(allxy).T
sx, sy = iqr_sigma(xx), iqr_sigma(yy)
sys.stderr.write("Estimated peak position variations %.3f %.3f\n" %
(sx, sy))
zcut = bonferroni(args.zcutoff, Np)
sys.stderr.write(
"Requested Z-score cutoff (%.2f) adjusted to %.2f (Bonferroni correction)\n"
% (args.zcutoff, zcut))
xcut, ycut = zcut * sx, zcut * sy
sys.stderr.write("Peak mismatch cutoffs set to %.3f / %.3f ppm\n" %
(xcut, ycut))
for (i, sample) in enumerate(eval(args.holonums)):
holopeaks = hps[i]
pms = holopeaks.peakmatch(peaks)
Nm, indm, xm, ym = holopeaks.matchcut_xy(peaks, xcut, ycut)
samplenum = str(sample).ljust(3)
print "Sample #" + samplenum + ": %d peaks shifted" % Nm
if args.print_peakmatch:
for (j, flag) in enumerate(indm):
if flag or args.print_full:
print samplenum + " %7.3f %7.3f %s %7.3f %7.3f" % (
holopeaks.peaks[pms[0][j]][0], holopeaks.peaks[
pms[0][j]][1], peaks.peaks[j][2], xm[j], ym[j])
def autocros(args, xargs):
apo = hsqc(os.path.join(args.folder, str(args.aponum) + ".nv"))
lrbt = [args.bleft, args.bright, args.bbottom, args.btop]
apo_peakdata = apo.peak_search(args.num_peaks, lrbt)
apo_peaks = peakset(apo_peakdata)
pmv = []
for sample in eval(args.controls):
ctrl_apo = hsqc(os.path.join(args.folder, str(sample) + ".nv"))
ctrl_peaks = peakset(ctrl_apo.peak_search(args.num_peaks, lrbt))
pmv.append(apo_peaks.matchsignal(ctrl_apo, True))
pmv = array(pmv)
apo_good = []
for (i, value) in enumerate(pmv.min(0)):
if value > args.vcutoff:
apo_good.append(apo_peakdata[i])
print "Found %d good peaks in set of %d" % (len(apo_good), 160)
good_peaks = peakset(array(apo_good))
for sample in eval(args.holonums):
holo = hsqc(os.path.join(args.folder, str(sample) + ".nv"))
x = good_peaks.matchsignal(holo, True)
print "Sample #" + str(sample).ljust(3) + ": " + " ".join(
map(lambda x: "%5.2f" % x,
map(lambda p: scoreatpercentile(x, p), range(0, 100 + 10,
10))))
def onkeypress(self, event):
PeakWindow.onkeypress(self, event)
if event.key == 'A':
self.set_apo_peaks(self.peaks)
print str(len(self.peaks)) + " apo-peaks stored"
if event.key == 'H':
self.load_current_holohsqc()
self.peak_search()
self.plot()
self.apoplot()
if event.key == 'N':
self.curholo = min(len(self.holonums) - 1, self.curholo + 1)
print "Now loading sample #" + str(self.holonums[self.curholo])
self.load_current_holohsqc()
self.peak_search()
self.plot()
self.apoplot()
print " ".join(
map(lambda x: "%5.2f" % x,
peakset(self.peaks).matchperc(peakset(self.apo_peaks))))
if event.key == 'P':
self.curholo = max(0, self.curholo - 1)
print "Now loading sample #" + str(self.holonums[self.curholo])
self.load_current_holohsqc()
self.peak_search()
self.plot()
self.apoplot()
print " ".join(
map(lambda x: "%5.2f" % x,
peakset(self.peaks).matchperc(peakset(self.apo_peaks))))
if event.key == ']':
self.acounter += 1
self.limit_acounter()
self.onapozoom(event)
if event.key == '[':
self.acounter -= 1
self.limit_acounter()
self.onapozoom(event)
if event.key == 'I':
apox = peakset(self.apo_peaks)
x = apox.matchsignal(self.dataset, True)
print "Sample #" + str(
self.holonums[self.curholo]).ljust(3) + ": " + " ".join(
map(
lambda x: "%5.2f" % x,
map(lambda p: scoreatpercentile(x, p),
range(0, 100 + 10, 10))))
if event.key == 'D':
apox = peakset(self.apo_peaks)
f = apox.matchsignal(self.dataset, True)
with open(str(self.holonums[self.curholo]) + ".out", 'w') as fout:
fout.write("RPH: " + " ".join(map(lambda x: "%6.3f" % x, f)) +
'\n')
if event.key == 'Q':
apox = peakset(self.apo_peaks)
with open("holall.dat", 'w') as fout:
for sample in self.holonums:
holo = hsqc(str(sample) + ".nv")
f = apox.matchsignal(holo, True)
fout.write("RPHI " + str(sample).ljust(3) + " :" +
" ".join(map(lambda x: "%6.3f" % x, f)) + '\n')
fout.write("RPHD " + str(sample).ljust(3) + " :" +
" ".join(
map(
lambda x: "%5.2f" % x,
map(lambda p: scoreatpercentile(f, p),
range(0, 100 + 10, 10)))) + '\n')
print "RPHD " + str(sample).ljust(3) + " :" + " ".join(
map(
lambda x: "%5.2f" % x,
map(lambda p: scoreatpercentile(f, p),
range(0, 100 + 10, 10))))