def plot_snp_dens_distrib(variations, window_size, data_dir, write_bg=False):
# Calculate and plot variations density distribution
density = calc_snp_density(variations, window_size)
density_distrib, bins = histogram(density, 20)
fpath = join(data_dir, 'snps_density.png')
title = 'SNP density distribution per {} bp windows'.format(window_size)
plot_distrib(density_distrib, bins, fhand=open(fpath, 'w'), color='c',
mpl_params={'set_xlabel': {'args': ['SNP density'],
'kwargs': {}},
'set_ylabel': {'args': ['SNP number'],
'kwargs': {}},
'set_title': {'args': [title], 'kwargs': {}},
'set_yscale': {'args': ['log'], 'kwargs': {}}})
# Manhattan plot for SNP density
fpath = join(data_dir, 'snps_density_manhattan.png')
fhand = open(fpath, 'w')
title = 'SNP denisity along the genome'
chrom = _load_matrix(variations, CHROM_FIELD)
pos = _load_matrix(variations, POS_FIELD)
manhattan_plot(chrom, pos, density,
mpl_params={'set_xlabel': {'args': ['Chromosome'],
'kwargs': {}},
'set_ylabel': {'args': ['SNP per {} bp'.format(window_size)],
'kwargs': {}},
'set_title': {'args': [title], 'kwargs': {}}},
fhand=fhand, figsize=(15, 7.5), ylim=1)
# Save in bedgraph format
if write_bg:
bg_fhand = open(join(data_dir, 'snp_density.bg'), 'w')
pos_dens = PositionalStatsCalculator(chrom, pos, density)
pos_dens.write(bg_fhand, 'snp_density',
'SNP number in {} bp around'.format(window_size),
track_type='bedgraph')
def test_manhattan_plot(self):
assert _look_for_first_different([1, 1, 3], 0) == 2
assert _look_for_first_different([1, 1, 3], 1) == 2
assert _look_for_first_different([1, 1, 3], 2) == 3
assert _look_for_first_different([1, 1, 1], 0) == 3
chrom = numpy.array([b"chr1"] * 3 + [b"chr2"] * 3 + [b"chr3"] * 3)
pos = numpy.array([1, 2, 3, 2, 5, 10, 1, 2, 3])
statistic = numpy.array([2, 3, 2, 5, 3, 1, 3, 4, 2])
with NamedTemporaryFile(suffix=".png") as fhand:
manhattan_plot(chrom, pos, statistic, fhand=fhand, figsize=(10, 10))
with NamedTemporaryFile(suffix=".png") as fhand:
manhattan_plot(chrom, pos, statistic, fhand=fhand, split_by_chrom=True)
def plot_inbreeding_coefficient(variations, max_num_allele, data_dir,
window_size, chunk_size=SNPS_PER_CHUNK,
min_num_genotypes=MIN_NUM_GENOTYPES_FOR_POP_STAT,
write_bg=False, calc_genome_wise=False):
# Calculate Inbreeding coefficient distribution
inbreed_coef = calc_inbreeding_coef(variations, chunk_size=chunk_size,
min_num_genotypes=min_num_genotypes)
ic_distrib, bins = histogram(inbreed_coef, 50, range_=(-1, 1))
fpath = join(data_dir, 'inbreeding_coef_distribution.png')
fhand = open(fpath, 'w')
title = 'Inbreeding coefficient distribution all samples'
plot_distrib(ic_distrib, bins, fhand=fhand,
mpl_params={'set_xlabel': {'args': ['Inbreeding coefficient'],
'kwargs': {}},
'set_ylabel': {'args': ['Number of SNPs'],
'kwargs': {}},
'set_title': {'args': [title], 'kwargs': {}},
'set_xlim': {'args': [-1, 1], 'kwargs': {}}})
# Save in bedgraph file
if calc_genome_wise:
bg_fhand = open(join(data_dir, 'ic.bg'), 'w')
chrom = _load_matrix(variations, CHROM_FIELD)
pos = _load_matrix(variations, POS_FIELD)
pos_ic = PositionalStatsCalculator(chrom, pos, inbreed_coef)
if write_bg:
pos_ic.write(bg_fhand, 'IC', 'Inbreeding coefficient',
track_type='bedgraph')
# Plot Ic along genome taking sliding windows
pos_ic = pos_ic.calc_window_stat()
chrom, pos, ic_windows = pos_ic.chrom, pos_ic.pos, pos_ic.stat
fpath = join(data_dir, 'ic_manhattan.png')
fhand = open(fpath, 'w')
title = 'Inbreeding coefficient (IC) along the genome'
manhattan_plot(chrom, pos, ic_windows, fhand=fhand, figsize=(15, 7.5),
ylim=-1,
mpl_params={'set_xlabel': {'args': ['Chromosome'],
'kwargs': {}},
'set_ylabel': {'args': ['IC'],
'kwargs': {}},
'set_title': {'args': [title], 'kwargs': {}}})
def plot_maf(variations, data_dir, chunk_size=SNPS_PER_CHUNK, window_size=None,
min_num_genotypes=MIN_NUM_GENOTYPES_FOR_POP_STAT, write_bg=False,
calc_genome_wise=False):
# Calculate and plot MAF distribution
mafs = calc_maf(variations, min_num_genotypes, chunk_size)
maf_distrib, bins = histogram(mafs, n_bins=25, range_=(0, 1))
fpath = join(data_dir, 'mafs.png')
title = 'Maximum allele frequency (MAF) distribution'
plot_distrib(maf_distrib, bins=bins, fhand=open(fpath, 'w'), color='c',
mpl_params={'set_xlabel': {'args': ['MAF'], 'kwargs': {}},
'set_ylabel': {'args': ['SNP number'],
'kwargs': {}},
'set_title': {'args': [title], 'kwargs': {}}})
# Write bedgraph file
if calc_genome_wise:
chrom = _load_matrix(variations, CHROM_FIELD)
pos = _load_matrix(variations, POS_FIELD)
bg_fhand = open(join(data_dir, 'maf.bg'), 'w')
pos_maf = PositionalStatsCalculator(chrom, pos, mafs,
window_size=window_size,
step=window_size)
if write_bg:
pos_maf.write(bg_fhand, 'MAF', 'Maximum allele frequency',
track_type='bedgraph')
if window_size is not None:
pos_maf = pos_maf.calc_window_stat()
# Manhattan plot for MAF along genome
fpath = join(data_dir, 'maf_manhattan.png')
fhand = open(fpath, 'w')
title = 'Max Allele Freq (MAF) along the genome'
chrom, pos, mafs = pos_maf.chrom, pos_maf.pos, pos_maf.stat
mpl_params = {'set_xlabel': {'args': ['Chromosome'], 'kwargs': {}},
'set_ylabel': {'args': ['MAF'],'kwargs': {}},
'set_title': {'args': [title], 'kwargs': {}}}
manhattan_plot(chrom, pos, mafs, mpl_params=mpl_params,
fhand=fhand, figsize=(15, 7.5))
def plot_r2(variations, window_size, data_dir, write_bg=False):
# Calculate LD r2 parameter in windows
chrom, pos, r2 = calc_r2_windows(variations, window_size=window_size)
# Plot r2 distribution
fpath = os.path.join(data_dir, 'r2_distrib.png')
distrib, bins = histogram(r2, n_bins=50, range_=(0, 1))
title = 'r2 distribution in windows of {} bp'.format(window_size)
mpl_params={'set_xlabel': {'args': ['r2'], 'kwargs': {}},
'set_ylabel': {'args': ['Number of windows'], 'kwargs': {}},
'set_title': {'args': [title], 'kwargs': {}}}
plot_distrib(distrib, bins, fhand=open(fpath, 'w'), figsize=(7, 7),
mpl_params=mpl_params)
# Manhattan plot
mask = numpy.logical_not(numpy.isnan(r2))
chrom = chrom[mask]
pos = pos[mask]
r2 = r2[mask]
fpath = os.path.join(data_dir, 'r2_manhattan.png')
title = 'r2 along genome in windows of {} bp'.format(window_size)
mpl_params={'set_xlabel': {'args': ['Chromosome'], 'kwargs': {}},
'set_ylabel': {'args': ['r2'], 'kwargs': {}},
'set_title': {'args': [title], 'kwargs': {}}}
manhattan_plot(chrom, pos, r2, fhand=open(fpath, 'w'), figsize=(15, 7),
marker='k', mpl_params=mpl_params)
# Write bg
if write_bg:
fpath = os.path.join(data_dir, 'r2_windows_{}.png'.format(window_size))
bg_fhand = open(fpath, 'w')
pos_r2 = PositionalStatsCalculator(chrom, pos, r2,
window_size=window_size,
step=window_size,
take_windows=False)
description = 'mean r2 in windows of {} bp'.format(window_size)
pos_r2.write(bg_fhand, 'r2', description, track_type='bedgraph')
def test_manhattan_plot(self):
assert _look_for_first_different([1, 1, 3], 0) == 2
assert _look_for_first_different([1, 1, 3], 1) == 2
assert _look_for_first_different([1, 1, 3], 2) == 3
assert _look_for_first_different([1, 1, 1], 0) == 3
chrom = numpy.array([b'chr1'] * 3 + [b'chr2'] * 3 + [b'chr3'] * 3)
pos = numpy.array([1, 2, 3, 2, 5, 10, 1, 2, 3])
statistic = numpy.array([2, 3, 2, 5, 3, 1, 3, 4, 2])
with NamedTemporaryFile(suffix='.png') as fhand:
manhattan_plot(chrom,
pos,
statistic,
fhand=fhand,
figsize=(10, 10))
with NamedTemporaryFile(suffix='.png') as fhand:
manhattan_plot(chrom,
pos,
statistic,
fhand=fhand,
split_by_chrom=True)
def plot_nucleotide_diversity_measures(variations, max_num_alleles,
window_size, data_dir,
chunk_size=SNPS_PER_CHUNK,
write_bg=False,
min_num_genotypes=MIN_NUM_GENOTYPES_FOR_POP_STAT):
fig = Figure(figsize=(20, 20))
canvas = FigureCanvas(fig)
marker = 'k'
chrom = _load_matrix(variations, CHROM_FIELD)
pos = _load_matrix(variations, POS_FIELD)
# Number of variable positions per bp
snp_density = PositionalStatsCalculator(chrom, pos,
numpy.ones(pos.shape),
window_size=window_size,
step=window_size)
snp_density = snp_density.calc_window_stat()
bg_fhand = open(join(data_dir, 'diversity_s.bg'), 'w')
if write_bg:
snp_density.write(bg_fhand, 's',
'SNP density in windows of {} bp'.format(window_size),
track_type='bedgraph')
axes = fig.add_subplot(311)
title = 'Nucleotide diversity measures averaged in windows of {} bp'
title = title.format(window_size)
mpl_params = {'set_title': {'args': [title], 'kwargs': {}},
'set_ylabel': {'args': ['SNPs number / bp'], 'kwargs': {}},
'set_ylim': {'args': [0, 1.2*numpy.max(snp_density.stat)],
'kwargs': {}}}
manhattan_plot(snp_density.chrom, snp_density.pos, snp_density.stat,
mpl_params=mpl_params, axes=axes, ylim=0, show_chroms=False,
marker=marker)
# Watterson estimator of nucleotide diversity
n_seqs = variations[GT_FIELD].shape[1] * variations[GT_FIELD].shape[2]
correction_factor = numpy.sum(1 / numpy.arange(1, n_seqs))
watterson = snp_density
watterson.stat = watterson.stat / correction_factor
bg_fhand = open(join(data_dir, 'diversity_s.bg'), 'w')
description = 'SNP density in windows of {} bp'.format(window_size)
if write_bg:
watterson.write(bg_fhand, 's', description, track_type='bedgraph')
axes = fig.add_subplot(312)
mpl_params={'set_ylabel': {'args': ['Watterson estimator'], 'kwargs': {}},
'set_ylim': {'args': [0, 1.2*numpy.max(watterson.stat)],
'kwargs': {}}}
manhattan_plot(watterson.chrom, watterson.pos, watterson.stat,
mpl_params=mpl_params, axes=axes, ylim=0, show_chroms=False,
marker=marker)
# Expected heterozygosity (Pi)
exp_het = calc_expected_het(variations, chunk_size=chunk_size,
min_num_genotypes=min_num_genotypes)
pi = PositionalStatsCalculator(chrom, pos, exp_het,
window_size=window_size, step=window_size)
pi = pi.calc_window_stat()
bg_fhand = open(join(data_dir, 'diversity_pi.bg'), 'w')
description = 'Pi in windows of {} bp'.format(window_size)
if write_bg:
pi.write(bg_fhand, 's', description, track_type='bedgraph')
axes = fig.add_subplot(313)
mpl_params={'set_xlabel': {'args': ['Chromosome'], 'kwargs': {}},
'set_ylabel': {'args': ['Pi'], 'kwargs': {}},
'set_ylim': {'args': [0, 1.2*numpy.max(pi.stat)],
'kwargs': {}}}
manhattan_plot(pi.chrom, pi.pos, pi.stat, axes=axes, ylim=0, marker=marker,
mpl_params=mpl_params)
canvas.print_figure(open(join(data_dir, 'nucleotide_diversity.png'), 'w'))
def plot_nucleotide_diversity_measures(
variations,
max_num_alleles,
window_size,
data_dir,
chunk_size=SNPS_PER_CHUNK,
write_bg=False,
min_num_genotypes=MIN_NUM_GENOTYPES_FOR_POP_STAT):
fig = Figure(figsize=(20, 20))
canvas = FigureCanvas(fig)
marker = 'k'
chrom = _load_matrix(variations, CHROM_FIELD)
pos = _load_matrix(variations, POS_FIELD)
# Number of variable positions per bp
snp_density = PositionalStatsCalculator(chrom,
pos,
numpy.ones(pos.shape),
window_size=window_size,
step=window_size)
snp_density = snp_density.calc_window_stat()
bg_fhand = open(join(data_dir, 'diversity_s.bg'), 'w')
if write_bg:
snp_density.write(
bg_fhand,
's',
'SNP density in windows of {} bp'.format(window_size),
track_type='bedgraph')
axes = fig.add_subplot(311)
title = 'Nucleotide diversity measures averaged in windows of {} bp'
title = title.format(window_size)
mpl_params = {
'set_title': {
'args': [title],
'kwargs': {}
},
'set_ylabel': {
'args': ['SNPs number / bp'],
'kwargs': {}
},
'set_ylim': {
'args': [0, 1.2 * numpy.max(snp_density.stat)],
'kwargs': {}
}
}
manhattan_plot(snp_density.chrom,
snp_density.pos,
snp_density.stat,
mpl_params=mpl_params,
axes=axes,
ylim=0,
show_chroms=False,
marker=marker)
# Watterson estimator of nucleotide diversity
n_seqs = variations[GT_FIELD].shape[1] * variations[GT_FIELD].shape[2]
correction_factor = numpy.sum(1 / numpy.arange(1, n_seqs))
watterson = snp_density
watterson.stat = watterson.stat / correction_factor
bg_fhand = open(join(data_dir, 'diversity_s.bg'), 'w')
description = 'SNP density in windows of {} bp'.format(window_size)
if write_bg:
watterson.write(bg_fhand, 's', description, track_type='bedgraph')
axes = fig.add_subplot(312)
mpl_params = {
'set_ylabel': {
'args': ['Watterson estimator'],
'kwargs': {}
},
'set_ylim': {
'args': [0, 1.2 * numpy.max(watterson.stat)],
'kwargs': {}
}
}
manhattan_plot(watterson.chrom,
watterson.pos,
watterson.stat,
mpl_params=mpl_params,
axes=axes,
ylim=0,
show_chroms=False,
marker=marker)
# Expected heterozygosity (Pi)
exp_het = calc_expected_het(variations,
chunk_size=chunk_size,
min_num_genotypes=min_num_genotypes)
pi = PositionalStatsCalculator(chrom,
pos,
exp_het,
window_size=window_size,
step=window_size)
pi = pi.calc_window_stat()
bg_fhand = open(join(data_dir, 'diversity_pi.bg'), 'w')
description = 'Pi in windows of {} bp'.format(window_size)
if write_bg:
pi.write(bg_fhand, 's', description, track_type='bedgraph')
axes = fig.add_subplot(313)
mpl_params = {
'set_xlabel': {
'args': ['Chromosome'],
'kwargs': {}
},
'set_ylabel': {
'args': ['Pi'],
'kwargs': {}
},
'set_ylim': {
'args': [0, 1.2 * numpy.max(pi.stat)],
'kwargs': {}
}
}
manhattan_plot(pi.chrom,
pi.pos,
pi.stat,
axes=axes,
ylim=0,
marker=marker,
mpl_params=mpl_params)
canvas.print_figure(open(join(data_dir, 'nucleotide_diversity.png'), 'w'))
def plot_inbreeding_coefficient(
variations,
max_num_allele,
data_dir,
window_size,
chunk_size=SNPS_PER_CHUNK,
min_num_genotypes=MIN_NUM_GENOTYPES_FOR_POP_STAT,
write_bg=False,
calc_genome_wise=False):
# Calculate Inbreeding coefficient distribution
inbreed_coef = calc_inbreeding_coef(variations,
chunk_size=chunk_size,
min_num_genotypes=min_num_genotypes)
ic_distrib, bins = histogram(inbreed_coef, 50, range_=(-1, 1))
fpath = join(data_dir, 'inbreeding_coef_distribution.png')
fhand = open(fpath, 'w')
title = 'Inbreeding coefficient distribution all samples'
plot_distrib(ic_distrib,
bins,
fhand=fhand,
mpl_params={
'set_xlabel': {
'args': ['Inbreeding coefficient'],
'kwargs': {}
},
'set_ylabel': {
'args': ['Number of SNPs'],
'kwargs': {}
},
'set_title': {
'args': [title],
'kwargs': {}
},
'set_xlim': {
'args': [-1, 1],
'kwargs': {}
}
})
# Save in bedgraph file
if calc_genome_wise:
bg_fhand = open(join(data_dir, 'ic.bg'), 'w')
chrom = _load_matrix(variations, CHROM_FIELD)
pos = _load_matrix(variations, POS_FIELD)
pos_ic = PositionalStatsCalculator(chrom, pos, inbreed_coef)
if write_bg:
pos_ic.write(bg_fhand,
'IC',
'Inbreeding coefficient',
track_type='bedgraph')
# Plot Ic along genome taking sliding windows
pos_ic = pos_ic.calc_window_stat()
chrom, pos, ic_windows = pos_ic.chrom, pos_ic.pos, pos_ic.stat
fpath = join(data_dir, 'ic_manhattan.png')
fhand = open(fpath, 'w')
title = 'Inbreeding coefficient (IC) along the genome'
manhattan_plot(chrom,
pos,
ic_windows,
fhand=fhand,
figsize=(15, 7.5),
ylim=-1,
mpl_params={
'set_xlabel': {
'args': ['Chromosome'],
'kwargs': {}
},
'set_ylabel': {
'args': ['IC'],
'kwargs': {}
},
'set_title': {
'args': [title],
'kwargs': {}
}
})
def plot_snp_dens_distrib(variations, window_size, data_dir, write_bg=False):
# Calculate and plot variations density distribution
density = calc_snp_density(variations, window_size)
density_distrib, bins = histogram(density, 20)
fpath = join(data_dir, 'snps_density.png')
title = 'SNP density distribution per {} bp windows'.format(window_size)
plot_distrib(density_distrib,
bins,
fhand=open(fpath, 'w'),
color='c',
mpl_params={
'set_xlabel': {
'args': ['SNP density'],
'kwargs': {}
},
'set_ylabel': {
'args': ['SNP number'],
'kwargs': {}
},
'set_title': {
'args': [title],
'kwargs': {}
},
'set_yscale': {
'args': ['log'],
'kwargs': {}
}
})
# Manhattan plot for SNP density
fpath = join(data_dir, 'snps_density_manhattan.png')
fhand = open(fpath, 'w')
title = 'SNP denisity along the genome'
chrom = _load_matrix(variations, CHROM_FIELD)
pos = _load_matrix(variations, POS_FIELD)
manhattan_plot(chrom,
pos,
density,
mpl_params={
'set_xlabel': {
'args': ['Chromosome'],
'kwargs': {}
},
'set_ylabel': {
'args': ['SNP per {} bp'.format(window_size)],
'kwargs': {}
},
'set_title': {
'args': [title],
'kwargs': {}
}
},
fhand=fhand,
figsize=(15, 7.5),
ylim=1)
# Save in bedgraph format
if write_bg:
bg_fhand = open(join(data_dir, 'snp_density.bg'), 'w')
pos_dens = PositionalStatsCalculator(chrom, pos, density)
pos_dens.write(bg_fhand,
'snp_density',
'SNP number in {} bp around'.format(window_size),
track_type='bedgraph')
def plot_maf(variations,
data_dir,
chunk_size=SNPS_PER_CHUNK,
window_size=None,
min_num_genotypes=MIN_NUM_GENOTYPES_FOR_POP_STAT,
write_bg=False,
calc_genome_wise=False):
# Calculate and plot MAF distribution
mafs = calc_maf(variations, min_num_genotypes, chunk_size)
maf_distrib, bins = histogram(mafs, n_bins=25, range_=(0, 1))
fpath = join(data_dir, 'mafs.png')
title = 'Maximum allele frequency (MAF) distribution'
plot_distrib(maf_distrib,
bins=bins,
fhand=open(fpath, 'w'),
color='c',
mpl_params={
'set_xlabel': {
'args': ['MAF'],
'kwargs': {}
},
'set_ylabel': {
'args': ['SNP number'],
'kwargs': {}
},
'set_title': {
'args': [title],
'kwargs': {}
}
})
# Write bedgraph file
if calc_genome_wise:
chrom = _load_matrix(variations, CHROM_FIELD)
pos = _load_matrix(variations, POS_FIELD)
bg_fhand = open(join(data_dir, 'maf.bg'), 'w')
pos_maf = PositionalStatsCalculator(chrom,
pos,
mafs,
window_size=window_size,
step=window_size)
if write_bg:
pos_maf.write(bg_fhand,
'MAF',
'Maximum allele frequency',
track_type='bedgraph')
if window_size is not None:
pos_maf = pos_maf.calc_window_stat()
# Manhattan plot for MAF along genome
fpath = join(data_dir, 'maf_manhattan.png')
fhand = open(fpath, 'w')
title = 'Max Allele Freq (MAF) along the genome'
chrom, pos, mafs = pos_maf.chrom, pos_maf.pos, pos_maf.stat
mpl_params = {
'set_xlabel': {
'args': ['Chromosome'],
'kwargs': {}
},
'set_ylabel': {
'args': ['MAF'],
'kwargs': {}
},
'set_title': {
'args': [title],
'kwargs': {}
}
}
manhattan_plot(chrom,
pos,
mafs,
mpl_params=mpl_params,
fhand=fhand,
figsize=(15, 7.5))
def plot_r2(variations, window_size, data_dir, write_bg=False):
# Calculate LD r2 parameter in windows
chrom, pos, r2 = calc_r2_windows(variations, window_size=window_size)
# Plot r2 distribution
fpath = os.path.join(data_dir, 'r2_distrib.png')
distrib, bins = histogram(r2, n_bins=50, range_=(0, 1))
title = 'r2 distribution in windows of {} bp'.format(window_size)
mpl_params = {
'set_xlabel': {
'args': ['r2'],
'kwargs': {}
},
'set_ylabel': {
'args': ['Number of windows'],
'kwargs': {}
},
'set_title': {
'args': [title],
'kwargs': {}
}
}
plot_distrib(distrib,
bins,
fhand=open(fpath, 'w'),
figsize=(7, 7),
mpl_params=mpl_params)
# Manhattan plot
mask = numpy.logical_not(numpy.isnan(r2))
chrom = chrom[mask]
pos = pos[mask]
r2 = r2[mask]
fpath = os.path.join(data_dir, 'r2_manhattan.png')
title = 'r2 along genome in windows of {} bp'.format(window_size)
mpl_params = {
'set_xlabel': {
'args': ['Chromosome'],
'kwargs': {}
},
'set_ylabel': {
'args': ['r2'],
'kwargs': {}
},
'set_title': {
'args': [title],
'kwargs': {}
}
}
manhattan_plot(chrom,
pos,
r2,
fhand=open(fpath, 'w'),
figsize=(15, 7),
marker='k',
mpl_params=mpl_params)
# Write bg
if write_bg:
fpath = os.path.join(data_dir, 'r2_windows_{}.png'.format(window_size))
bg_fhand = open(fpath, 'w')
pos_r2 = PositionalStatsCalculator(chrom,
pos,
r2,
window_size=window_size,
step=window_size,
take_windows=False)
description = 'mean r2 in windows of {} bp'.format(window_size)
pos_r2.write(bg_fhand, 'r2', description, track_type='bedgraph')
