def create_ols_legend_plot(matplot_dict, motif_freq,
sample_scores, control_scores,
flank_length, legend_title):
f = matplot_dict['figure']
gs = matplot_dict['gridspec']
phlyop_plots_legend = plt.Subplot(f, gs, autoscale_on=True)
corr_result = get_pearson_corr(motif_freq,
remove_flanking_scores(sample_scores, flank_length))
corr_pval = corr_result[1]
corr_r2 = corr_result[0]
ttest_result = perform_t_test(remove_flanking_scores(sample_scores, flank_length),
get_flanking_scores(sample_scores, flank_length))
p_deltaphylop = ttest_result['one_sided_pval']
delta_phylop = ttest_result['delta']
#T_deltaphylop = ttest_result['T']
pearsonr_pval = str('%.1g'%corr_pval)
if 'e' in pearsonr_pval:
pearsonr_pval += '}'
pearsonr_pval = pearsonr_pval.replace('e', '*10^{').replace('-0','-')
score_pval = str('%.1g'%p_deltaphylop)
if 'e' in score_pval:
score_pval += '}'
score_pval = score_pval.replace('e', '*10^{').replace('-0','-')
textstr = r'$r^2_{pearson}=%.2f(p=%s)$' '\n' r'$\Delta_{%s}=%.2f(p=%s)$' %(corr_r2, pearsonr_pval, legend_title, delta_phylop, score_pval)
txtx = 1-LEGEND_XMULTIPLIER*len(textstr)/100.0
phlyop_plots_legend.set_frame_on(False)
phlyop_plots_legend.set_xticks([])
phlyop_plots_legend.set_yticks([])
phlyop_plots_legend.text(txtx, TXT_YPOS, textstr, fontsize=LEGEND_FONTSIZE)
f.add_subplot(phlyop_plots_legend)
def create_scatter_plot(matplot_dict, motif_freq,
sample_scores, control_scores,
flank_length, num_occurrences, y_label):
f = matplot_dict['figure']
gs = matplot_dict['gridspec']
scatter_plot = plt.Subplot(f, gs, autoscale_on=True)
control_scores = remove_flanking_scores(control_scores, flank_length)
sample_scores = remove_flanking_scores(sample_scores, flank_length)
fit = np.polyfit(motif_freq, sample_scores, 1)
fit_fn = np.poly1d(fit)
control_ols = perform_OLS(control_scores, motif_freq)
sample_ols = perform_OLS(sample_scores, motif_freq)
sample_regression_line = sample_ols['regression_line']
control_regression_line = control_ols['regression_line']
s1 = scatter_plot.scatter(motif_freq, sample_scores, color='g',
s=[POINTSIZE for i in motif_freq],
marker='^', label=r'$\mathrm{Sample}$')
scatter_plot.plot(motif_freq, sample_regression_line, 'g',
motif_freq, fit_fn(motif_freq),
color='g', linewidth=LINEWIDTH)
s2 = scatter_plot.scatter(motif_freq, control_scores,
color=GREYNESS, s=[POINTSIZE for i in motif_freq],
marker='o', label=r'$\mathrm{Control}$')
scatter_plot.plot(motif_freq, control_regression_line,
color=GREYNESS, linewidth=LINEWIDTH)
leg = scatter_plot.legend(fontsize=14)
leg.draw_frame(True)
#leg.get_frame().set_edgecolor('b')
leg.get_frame().set_linewidth(2.0)
ticks_and_labels = np.linspace(1.02*min(motif_freq), 1.02*max(motif_freq),
num = 3, endpoint=True)
scatter_plot.set_xticks(ticks_and_labels)
ticks_and_labels = ["$%.2f$"%(x/(1.02*num_occurrences)) for x in ticks_and_labels]
scatter_plot.set_xticklabels(ticks_and_labels)#, rotation=45)
yloc = plt.MaxNLocator(MAX_YTICKS)
scatter_plot.yaxis.set_major_locator(yloc)
scatter_plot.set_xlabel(r'$\mathrm{Most}\ \mathrm{frequent} \ \mathrm{base}\ \mathrm{frequency}$',
fontsize=FONTSIZE, fontweight='bold')
scatter_plot.get_xaxis().tick_bottom()
scatter_plot.get_yaxis().tick_left()
scatter_plot.set_ylabel('$\mathrm{%s}\ \mathrm{Score}$'%(y_label), fontsize=FONTSIZE, fontweight='bold')
scatter_plot.tick_params(axis='y', which='major', pad=TICKPAD)
scatter_plot.tick_params(axis='x', which='major', pad=TICKPAD)
scatter_plot.get_yaxis().set_tick_params(direction='out')
scatter_plot.get_xaxis().set_tick_params(direction='out')
scatter_plot.tick_params('both', length=TICKLENGTH, width=2, which='major')
f.add_subplot(scatter_plot)
'motif_enrichment': motif_enrichment,
'motif_logo': motif_logo_encoded,
'motif_logorc': motif_logorc_encoded,
'moca_plot': moca_plot_encoded,
'moca_plotrc': moca_plotrc_encoded,
'no_fimo_hit_sample': True
})
continue
gerp_mean_sample = np.loadtxt(
os.path.join(fimo_sample, 'gerp.mean.txt')).tolist()
phylop_mean_sample = np.loadtxt(
os.path.join(fimo_sample, 'phylop.mean.txt')).tolist()
delta_phylop_ttest = perform_t_test(
remove_flanking_scores(phylop_mean_sample, flank_length),
get_flanking_scores(phylop_mean_sample, flank_length))
p_delta_phylop = delta_phylop_ttest['one_sided_pval']
delta_phylop = delta_phylop_ttest['delta']
delta_gerp_ttest = perform_t_test(
remove_flanking_scores(gerp_mean_sample, flank_length),
get_flanking_scores(gerp_mean_sample, flank_length))
p_delta_gerp = delta_gerp_ttest['one_sided_pval']
delta_gerp = delta_gerp_ttest['delta']
phylop_sample_ols = perform_OLS(
remove_flanking_scores(phylop_mean_sample, flank_length),
motif_freq)
gerp_sample_ols = perform_OLS(
remove_flanking_scores(gerp_mean_sample, flank_length), motif_freq)
motif_evalue = get_motif_evalue(record)
if os.stat(os.path.join(fimo_sample, 'gerp.mean.txt')).st_size == 0:
db.encode_tf_stats.insert_one({ 'encode_id': d,
'motif_number': i+1,
'center_enrichment': center_enrichment,
'center_enrichment_pval': center_enrichment_pval,
'motif_evalue': motif_evalue,
'motif_enrichment': motif_enrichment,
'no_fimo_hit_sample': True})
continue
gerp_mean_sample = np.loadtxt(os.path.join(fimo_sample, 'gerp.mean.txt')).tolist()
phylop_mean_sample = np.loadtxt(os.path.join(fimo_sample, 'phylop.mean.txt')).tolist()
delta_phylop_ttest = perform_t_test(remove_flanking_scores(phylop_mean_sample, flank_length),
get_flanking_scores(phylop_mean_sample, flank_length))
p_delta_phylop = delta_phylop_ttest['one_sided_pval']
delta_phylop = delta_phylop_ttest['delta']
delta_gerp_ttest = perform_t_test(remove_flanking_scores(gerp_mean_sample, flank_length),
get_flanking_scores(gerp_mean_sample, flank_length))
p_delta_gerp = delta_gerp_ttest['one_sided_pval']
delta_gerp = delta_gerp_ttest['delta']
phylop_sample_ols = perform_OLS(remove_flanking_scores(phylop_mean_sample, flank_length), motif_freq)
gerp_sample_ols = perform_OLS(remove_flanking_scores(gerp_mean_sample, flank_length), motif_freq)
phylop_sample_fit = phylop_sample_ols['regression_fit']
gerp_sample_fit = gerp_sample_ols['regression_fit']
corr_phylop_sample = get_pearson_corr(motif_freq, remove_flanking_scores(phylop_mean_sample, flank_length))
corr_gerp_sample = get_pearson_corr(motif_freq, remove_flanking_scores(gerp_mean_sample, flank_length))
