def plots(input_dir, date, data_filter, virus, passage_order, transition_order, pairs, label_order, pairs_adar, filter_reads=None):
output_dir = input_dir + date + "_plots"
plus_minus = u"\u00B1"
try:
os.mkdir(output_dir)
except OSError:
print("Creation of the directory %s failed" % output_dir)
else:
print("Successfully created the directory %s " % output_dir)
if filter_reads is True:
data_filter["no_variants"] = np.where(data_filter["Prob"] < 0.95, 0, data_filter["no_variants"])
data_filter["Read_count"] = data_filter[data_filter["Read_count"] > 10000]
mutation_order = ["A>G", "U>C", "G>A", "C>U", "A>C", "U>G", "A>U", "U>A", "G>C", "C>G", "C>A", "G>U"]
type_order = ["Synonymous", "Non-Synonymous", "Premature Stop Codon"]
# g1 = sns.catplot("label", "frac_and_weight", data=data_filter, hue="Mutation", order=label_order, palette="tab20",
# kind="point", dodge=True, hue_order=mutation_order, join=False, estimator=weighted_varaint,
# orient="v")
# g1.set_axis_labels("Passage", "Variant Frequency {} CI=95%".format(plus_minus))
# g1.set_xticklabels(fontsize=9, rotation=45)
# g1.set(yscale='log')
# g1.set(ylim=(10**-5, 10**-1))
#
# # plt.show()
# g1.savefig(output_dir + "/All_Mutations_point_plot", dpi=300)
# plt.close()
data_filter["passage"] = data_filter["passage"].astype(str)
data_filter["passage"] = np.where(data_filter["passage"] != "RNA\nControl", "p" + data_filter["passage"], data_filter["passage"])
g2 = sns.catplot("passage", "frac_and_weight", data=data_filter, hue="Mutation", order=passage_order,
palette=mutation_palette(4)
, kind="point", dodge=0.5, hue_order=transition_order, join=False, estimator=weighted_varaint,
orient="v")
g2.set_axis_labels("Passage", "Variant Frequency {} CI=95%".format(plus_minus))
g2.set(yscale='log')
g2.set(ylim=(10 ** -6, 10 ** -2))
# g2.set_xticklabels(fontsize=10, rotation=45)
# g2.savefig("/Users/odedkushnir/Google Drive/Studies/PhD/Prgress reports/20200913 Final report/plots" +
# "/Transition_Mutations_point_plot_Mahoney", dpi=300)
g2.savefig(output_dir + "/Transition_Mutations_point_plot_{0}".format(virus), dpi=300)
plt.close()
passage_g = sns.boxplot(x="passage", y="Frequency", data=data_filter, hue="Mutation", order=passage_order,
palette=mutation_palette(4), dodge=True, hue_order=transition_order)
passage_g.set_yscale('log')
passage_g.set_ylim(10 ** -6, 10 ** -1)
passage_g.set(xlabel="Passage", ylabel="Variant Frequency")
annot = Annotator(passage_g, pairs, x="passage", y="Frequency", hue="Mutation", data=data_filter,
order=passage_order, hue_order=transition_order)
annot.configure(test='t-test_welch', text_format='star', loc='outside', verbose=2,
comparisons_correction="Bonferroni")
annot.apply_test()
file_path = output_dir + "/sts.csv"
with open(file_path, "w") as o:
with contextlib.redirect_stdout(o):
passage_g, test_results = annot.annotate()
plt.legend(bbox_to_anchor=(1.05, 0.5), loc=2, borderaxespad=0.)
plt.tight_layout()
plt.savefig(output_dir + "/Transition_Mutations_box_stat_plot_{0}".format(virus), dpi=300)
plt.close()
data_filter_synonymous = data_filter.loc[data_filter.Type == "Synonymous"]
data_filter_synonymous["Mutation"] = np.where(((data_filter_synonymous["Mutation"] == "A>G") &
(data_filter_synonymous["5`_ADAR_Preference"] == "High")),
"High\nADAR-like\nA>G", np.where(((data_filter_synonymous["Mutation"] == "A>G")
& (data_filter_synonymous["5`_ADAR_Preference"] == "Intermediate")),
"Intermediate\nADAR-like\nA>G",
np.where(((data_filter_synonymous["Mutation"] == "A>G") &
(data_filter_synonymous["5`_ADAR_Preference"] == "Low")),
"Low\nADAR-like\nA>G",
data_filter_synonymous["Mutation"])))
data_filter_synonymous["Mutation_adar"] = np.where(((data_filter_synonymous["Mutation"] == "U>C") &
(data_filter_synonymous["3`_ADAR_Preference"] == "High")),
"High\nADAR-like\nU>C", np.where(((data_filter_synonymous["Mutation"] == "U>C")
& (data_filter_synonymous["3`_ADAR_Preference"] == "Intermediate")),
"Intermediate\nADAR-like\nU>C",
np.where(((data_filter_synonymous["Mutation"] == "U>C") &
(data_filter_synonymous["3`_ADAR_Preference"] == "Low")),
"Low\nADAR-like\nU>C",
data_filter_synonymous["Mutation"])))
mutation_adar_order = ["High\nADAR-like\nA>G", "Low\nADAR-like\nA>G",
"High\nADAR-like\nU>C", "Low\nADAR-like\nU>C"]
data_filter_synonymous["passage"] = data_filter_synonymous["passage"].astype(str)
catplot_adar = sns.catplot(x="passage", y="frac_and_weight", data=data_filter_synonymous, hue="Mutation_adar",
order=passage_order, palette=mutation_palette(4, adar=True), kind="point", dodge=0.5,
hue_order=mutation_adar_order, join=False, estimator=weighted_varaint, orient="v",
legend=True)
catplot_adar.set_axis_labels("Passage", "Variant Frequency {0} CI=95%".format(plus_minus))
catplot_adar.set(yscale='log')
catplot_adar.set(ylim=(10 ** -6, 10 ** -2))
plt.savefig(output_dir + "/adar_pref_mutation_point_plot_{0}.png".format(virus), dpi=300)
plt.close()
adar_g = sns.boxplot(x="passage", y="Frequency", data=data_filter_synonymous, hue="Mutation_adar",
order=passage_order, palette=mutation_palette(4, adar=True), dodge=True,
hue_order=mutation_adar_order)
adar_g.set_yscale('log')
adar_g.set_ylim(10 ** -6, 10 ** -1)
adar_g.set(xlabel="Passage", ylabel="Variant Frequency")
annot = Annotator(adar_g, pairs_adar, x="passage", y="Frequency", hue="Mutation_adar",
data=data_filter_synonymous, hue_order=mutation_adar_order, order=passage_order)
annot.configure(test='t-test_welch', text_format='star', loc='outside', verbose=2,
comparisons_correction="Bonferroni")
annot.apply_test()
file_path = output_dir + "/sts_adar.csv"
with open(file_path, "w") as o:
with contextlib.redirect_stdout(o):
adar_g, test_results = annot.annotate()
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)
plt.tight_layout()
plt.savefig(output_dir + "/adar_pref_mutation_box_plot_{0}.png".format(virus), dpi=300)
plt.close()
def main():
replica_lst = (1, 2, 3)
input_dir = "/Volumes/STERNADILABHOME$/volume3/okushnir/AccuNGS/20201008RV-202329127/merged/passages/"
prefix = "inosine_predict_context"
output_dir = input_dir + "20201112_10000coverage_%s" % prefix
try:
os.mkdir(output_dir)
except OSError:
print("Creation of the directory %s failed" % output_dir)
else:
print("Successfully created the directory %s " % output_dir)
for replica in replica_lst:
data_filter = pd.read_pickle(input_dir + prefix + "/data_filter.pkl")
data_filter_ag = pd.read_pickle(input_dir + prefix +
"/data_filter_ag.pkl")
data_filter_uc = pd.read_pickle(input_dir + prefix +
"/data_filter_uc.pkl")
data_filter["passage"] = data_filter["passage"].astype(int)
data_filter_ag["passage"] = data_filter_ag["passage"].astype(int)
data_filter_uc["passage"] = data_filter_uc["passage"].astype(int)
data_filter = data_filter[(data_filter['Read_count'] > 10000)]
data_filter_ag = data_filter_ag[(data_filter_ag['Read_count'] > 10000)]
data_filter_uc = data_filter_uc[(data_filter_uc['Read_count'] > 10000)]
data_filter = data_filter[data_filter["label"] != "p10-1"]
data_filter_ag = data_filter_ag[data_filter_ag["label"] != "p10-1"]
data_filter_uc = data_filter_uc[data_filter_uc["label"] != "p10-1"]
data_filter["replica"] = np.where(
data_filter["label"] == "RNA Control\nPrimer ID", replica,
data_filter["replica"])
data_filter_ag["replica"] = np.where(
data_filter_ag["label"] == "RNA Control\nPrimer ID", replica,
data_filter_ag["replica"])
data_filter_uc["replica"] = np.where(
data_filter_uc["label"] == "RNA Control\nPrimer ID", replica,
data_filter_uc["replica"])
data_filter = data_filter[data_filter["replica"] == replica]
data_filter_ag = data_filter_ag[data_filter_ag["replica"] == replica]
data_filter_uc = data_filter_uc[data_filter_uc["replica"] == replica]
#Plots
transition_order = ["A>G", "U>C", "G>A", "C>U"]
type_order = ["Synonymous", "Non-Synonymous", "Premature Stop Codon"]
type_order_ag = ["Synonymous", "Non-Synonymous"]
adar_preference = ["High", "Intermediate", "Low"]
data_filter_grouped = data_filter.groupby(
["label", "passage", "replica", "Type",
"Mutation"])["frac_and_weight"].agg(lambda x: weighted_varaint(x))
data_filter_grouped = data_filter_grouped.reset_index()
data_filter_grouped = data_filter_grouped.rename(
columns={"frac_and_weight": "Frequency"})
data_filter_grouped["Frequency"] = data_filter_grouped[
"Frequency"].astype(float)
data_filter_grouped = data_filter_grouped[
data_filter_grouped["label"] != "RNA Control\nPrimer ID"]
data_filter_grouped = data_filter_grouped[
data_filter_grouped["label"] != "RNA Control\nRND"]
data_filter_grouped["replica"] = data_filter_grouped["replica"].astype(
int)
data_filter_grouped = data_filter_grouped[
data_filter_grouped["replica"] == replica]
print(data_filter_grouped.to_string())
data_reg_ag = data_filter_grouped[data_filter_grouped["Mutation"] ==
"A>G"]
data_reg_uc = data_filter_grouped[data_filter_grouped["Mutation"] ==
"U>C"]
data_reg_ga = data_filter_grouped[data_filter_grouped["Mutation"] ==
"G>A"]
data_reg_cu = data_filter_grouped[data_filter_grouped["Mutation"] ==
"C>U"]
data_reg_ag_syn = data_reg_ag[data_reg_ag["Type"] == "Synonymous"]
data_reg_uc_syn = data_reg_uc[data_reg_uc["Type"] == "Synonymous"]
data_reg_ga_syn = data_reg_ga[data_reg_ga["Type"] == "Synonymous"]
data_reg_cu_syn = data_reg_cu[data_reg_cu["Type"] == "Synonymous"]
data_reg_ag_non_syn = data_reg_ag[data_reg_ag["Type"] ==
"Non-Synonymous"]
data_reg_uc_non_syn = data_reg_uc[data_reg_uc["Type"] ==
"Non-Synonymous"]
data_reg_ga_non_syn = data_reg_ga[data_reg_ga["Type"] ==
"Non-Synonymous"]
data_reg_cu_non_syn = data_reg_cu[data_reg_cu["Type"] ==
"Non-Synonymous"]
data_reg_ga_pmsc = data_reg_ga[data_reg_ga["Type"] ==
"Premature Stop Codon"]
data_reg_cu_pmsc = data_reg_cu[data_reg_cu["Type"] ==
"Premature Stop Codon"]
stat_slope1, stat_intercept1, r_value1, p_value1, std_err1 = stats.linregress(
data_reg_ag_syn['passage'], data_reg_ag_syn['Frequency'])
stat_slope2, stat_intercept2, r_value2, p_value2, std_err2 = stats.linregress(
data_reg_uc_syn['passage'], data_reg_uc_syn['Frequency'])
stat_slope3, stat_intercept3, r_value3, p_value3, std_err3 = stats.linregress(
data_reg_ga_syn['passage'], data_reg_ga_syn['Frequency'])
stat_slope4, stat_intercept4, r_value4, p_value4, std_err4 = stats.linregress(
data_reg_cu_syn['passage'], data_reg_cu_syn['Frequency'])
# data_reg_adar_nonsyn
stat_slope5, stat_intercept5, r_value5, p_value5, std_err5 = stats.linregress(
data_reg_ag_non_syn['passage'], data_reg_ag_non_syn['Frequency'])
stat_slope6, stat_intercept6, r_value6, p_value6, std_err6 = stats.linregress(
data_reg_uc_non_syn['passage'], data_reg_uc_non_syn['Frequency'])
stat_slope7, stat_intercept7, r_value7, p_value7, std_err7 = stats.linregress(
data_reg_ga_non_syn['passage'], data_reg_ga_non_syn['Frequency'])
stat_slope8, stat_intercept8, r_value8, p_value8, std_err8 = stats.linregress(
data_reg_cu_non_syn['passage'], data_reg_cu_non_syn['Frequency'])
#pmsc
stat_slope11, stat_intercept11, r_value11, p_value11, std_err11 = stats.linregress(
data_reg_ga_pmsc['passage'], data_reg_ga_pmsc['Frequency'])
stat_slope12, stat_intercept12, r_value12, p_value12, std_err12 = stats.linregress(
data_reg_cu_pmsc['passage'], data_reg_cu_pmsc['Frequency'])
data_filter_grouped = data_filter_grouped.rename(
columns={"passage": "Passage"})
reg_plot = sns.lmplot(x="Passage",
y="Frequency",
data=data_filter_grouped,
hue="Mutation",
hue_order=transition_order,
fit_reg=True,
col="Mutation",
col_order=transition_order,
row="Type",
row_order=type_order,
palette=mutation_palette(4),
line_kws={'label': "Linear Reg"},
legend=True,
height=6,
x_estimator=np.mean,
x_jitter=.05) # markers=["o", "v", "x"]
reg_plot.fig.subplots_adjust(wspace=.02)
ax = reg_plot.axes[0, 0]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
label_line_1 = "y={0:.3g}x+{1:.3g}".format(stat_slope1,
stat_intercept1)
label_line_2 = "y={0:.3g}x+{1:.3g}".format(stat_slope2,
stat_intercept2)
label_line_3 = "y={0:.3g}x+{1:.3g}".format(stat_slope3,
stat_intercept3)
label_line_4 = "y={0:.3g}x+{1:.3g}".format(stat_slope4,
stat_intercept4)
label_line_5 = "y={0:.3g}x+{1:.3g}".format(stat_slope5,
stat_intercept5)
label_line_6 = "y={0:.3g}x+{1:.3g}".format(stat_slope6,
stat_intercept6)
label_line_7 = "y={0:.3g}x+{1:.3g}".format(stat_slope7,
stat_intercept7)
label_line_8 = "y={0:.3g}x+{1:.3g}".format(stat_slope8,
stat_intercept8)
label_line_11 = "y={0:.3g}x+{1:.3g}".format(stat_slope11,
stat_intercept11)
label_line_12 = "y={0:.3g}x+{1:.3g}".format(stat_slope12,
stat_intercept12)
L_labels[0].set_text(label_line_1)
ax = reg_plot.axes[0, 1]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_2)
ax = reg_plot.axes[0, 2]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_3)
ax = reg_plot.axes[0, 3]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_4)
ax = reg_plot.axes[1, 0]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_5)
ax = reg_plot.axes[1, 1]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_6)
ax = reg_plot.axes[1, 2]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_7)
ax = reg_plot.axes[1, 3]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_8)
ax = reg_plot.axes[2, 2]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_11)
ax = reg_plot.axes[2, 3]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_12)
reg_plot.set(xlim=(0, 13))
reg_plot.set(ylim=(0.000, 0.001))
# reg_plot.fig.suptitle("RV #%s" % str(replica), y=0.99)
# plt.tight_layout()
reg_plot.savefig(output_dir +
"/transition_lmplot_replica%s.png" % str(replica),
dpi=300)
plt.close()
data_filter["Transition"] = data_filter.Mutation.str.contains("A>G") | data_filter.Mutation.str.contains("U>C") \
| data_filter.Mutation.str.contains("C>U") | data_filter.Mutation.str.contains("G>A")
data_transition = data_filter[data_filter["Transition"] == True]
position_mutation = sns.relplot(x="Pos",
y="Frequency",
data=data_transition,
hue="Mutation",
col="passage",
hue_order=transition_order,
col_wrap=3,
palette=mutation_palette(4),
height=4)
#, estimator=weighted_varaint)#,
position_mutation.set_axis_labels("", "Variant Frequency")
position_mutation.axes.flat[0].set_yscale('symlog', linthreshy=10**-5)
position_mutation.axes.flat[0].set_ylim(10**-6, 10**-1)
plt.savefig(output_dir + "/position_mutation_replica%s.png" % replica,
dpi=300)
plt.close()
columns = ["Mutation", "Replica", "Type", "Slope", "Intercept"]
mutation_rate_df = pd.DataFrame(columns=columns)
mutation_rate_df.loc[0] = [
"A>G", replica, "Synonymous", stat_slope1, stat_intercept1
]
mutation_rate_df.loc[1] = [
"U>C", replica, "Synonymous", stat_slope2, stat_intercept2
]
mutation_rate_df.loc[2] = [
"G>A", replica, "Synonymous", stat_slope3, stat_intercept3
]
mutation_rate_df.loc[3] = [
"C>U", replica, "Synonymous", stat_slope4, stat_intercept4
]
mutation_rate_df.loc[4] = [
"A>G", replica, "Non-Synonymous", stat_slope5, stat_intercept5
]
mutation_rate_df.loc[5] = [
"U>C", replica, "Non-Synonymous", stat_slope6, stat_intercept6
]
mutation_rate_df.loc[6] = [
"G>A", replica, "Non-Synonymous", stat_slope7, stat_intercept7
]
mutation_rate_df.loc[7] = [
"C>U", replica, "Non-Synonymous", stat_slope8, stat_intercept8
]
mutation_rate_df.loc[8] = [
"G>A", replica, "Pre Mature Stop Codon", stat_slope11,
stat_intercept11
]
mutation_rate_df.loc[9] = [
"C>U", replica, "Pre Mature Stop Codon", stat_slope12,
stat_intercept12
]
mutation_rate_df.to_csv(output_dir + "/mutation_rate%s.csv" % replica,
sep=',',
encoding='utf-8')
mutation_rate_df.to_pickle(output_dir +
"/mutation_rate%s.pkl" % replica)
# # A>G Prev Context
mutation_ag = sns.catplot("passage",
"frac_and_weight",
data=data_filter_ag,
hue="5`_ADAR_Preference",
palette=mutation_palette(3,
adar=True,
ag=True),
kind="point",
dodge=True,
estimator=weighted_varaint,
orient="v",
col="Type",
join=False,
col_order=type_order_ag,
hue_order=adar_preference)
mutation_ag.set(yscale="log")
mutation_ag.set(ylim=(1 * 10**-5, 1 * 10**-2))
mutation_ag.fig.suptitle("A>G ADAR_like Mutation in RV #%s" % replica,
y=0.99)
plt.subplots_adjust(top=0.85)
mutation_ag.set_axis_labels("Passage", "Variant Frequency")
mutation_ag.savefig(
output_dir + "/ag_ADAR_like_Mutation_col_replica%s.png" % replica,
dpi=300)
plt.close()
data_filter_ag_grouped = data_filter_ag.groupby([
"label", "Type", "passage", "replica", "ADAR_grade_five",
"5`_ADAR_Preference"
])["frac_and_weight"].agg(lambda x: weighted_varaint(x))
data_filter_ag_grouped = data_filter_ag_grouped.reset_index()
# print(data_filter_ag_grouped.to_string())
data_filter_ag_grouped = data_filter_ag_grouped.rename(
columns={"frac_and_weight": "Frequency"})
data_filter_ag_grouped["Frequency"] = data_filter_ag_grouped[
"Frequency"].astype(float)
data_filter_ag_grouped = data_filter_ag_grouped[
data_filter_ag_grouped["label"] != "RNA Control\nPrimer ID"]
data_filter_ag_grouped = data_filter_ag_grouped[
data_filter_ag_grouped["label"] != "RNA Control\nRND"]
data_filter_ag_grouped = data_filter_ag_grouped[
data_filter_ag_grouped["replica"] == replica]
data_reg_full_adar = data_filter_ag_grouped[
data_filter_ag_grouped["ADAR_grade_five"] == 1]
data_reg_semi_adar = data_filter_ag_grouped[
data_filter_ag_grouped["ADAR_grade_five"] == 0.5]
data_reg_nonadar = data_filter_ag_grouped[
data_filter_ag_grouped["ADAR_grade_five"] == 0]
data_reg_full_adar_syn = data_reg_full_adar[data_reg_full_adar["Type"]
== "Synonymous"]
data_reg_semi_adar_syn = data_reg_semi_adar[data_reg_semi_adar["Type"]
== "Synonymous"]
data_reg_nonadar_syn = data_reg_nonadar[data_reg_nonadar["Type"] ==
"Synonymous"]
data_reg_full_adar_non_syn = data_reg_full_adar[
data_reg_full_adar["Type"] == "Non-Synonymous"]
data_reg_semi_adar_non_syn = data_reg_semi_adar[
data_reg_semi_adar["Type"] == "Non-Synonymous"]
data_reg_nonadar_non_syn = data_reg_nonadar[data_reg_nonadar["Type"] ==
"Non-Synonymous"]
stat_slope1, stat_intercept1, r_value1, p_value1, std_err1 = stats.linregress(
data_reg_full_adar_syn['passage'],
data_reg_full_adar_syn['Frequency'])
stat_slope2, stat_intercept2, r_value2, p_value2, std_err2 = stats.linregress(
data_reg_semi_adar_syn['passage'],
data_reg_semi_adar_syn['Frequency'])
stat_slope3, stat_intercept3, r_value3, p_value3, std_err3 = stats.linregress(
data_reg_nonadar_syn['passage'], data_reg_nonadar_syn['Frequency'])
# data_reg_adar_nonsyn
stat_slope4, stat_intercept4, r_value4, p_value4, std_err4 = stats.linregress(
data_reg_full_adar_non_syn['passage'],
data_reg_full_adar_non_syn['Frequency'])
stat_slope5, stat_intercept5, r_value5, p_value5, std_err5 = stats.linregress(
data_reg_semi_adar_non_syn['passage'],
data_reg_semi_adar_non_syn['Frequency'])
stat_slope6, stat_intercept6, r_value6, p_value6, std_err6 = stats.linregress(
data_reg_nonadar_non_syn['passage'],
data_reg_nonadar_non_syn['Frequency'])
data_filter_ag_grouped = data_filter_ag_grouped.rename(
columns={"passage": "Passage"})
ag_reg_plot = sns.lmplot(x="Passage",
y="Frequency",
data=data_filter_ag_grouped,
hue="5`_ADAR_Preference",
hue_order=adar_preference,
markers=["o", "v", "x"],
fit_reg=True,
col="5`_ADAR_Preference",
col_order=adar_preference,
row="Type",
row_order=type_order_ag,
palette=mutation_palette(3,
adar=True,
ag=True),
line_kws={'label': "Linear Reg"},
legend=True,
height=6,
x_estimator=np.mean,
x_jitter=.05)
ag_reg_plot.fig.subplots_adjust(wspace=.02)
ax = ag_reg_plot.axes[0, 0]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
label_line_1 = "y={0:.3g}x+{1:.3g}".format(stat_slope1,
stat_intercept1)
label_line_2 = "y={0:.3g}x+{1:.3g}".format(stat_slope2,
stat_intercept2)
label_line_3 = "y={0:.3g}x+{1:.3g}".format(stat_slope3,
stat_intercept3)
label_line_4 = "y={0:.3g}x+{1:.3g}".format(stat_slope4,
stat_intercept4)
label_line_5 = "y={0:.3g}x+{1:.3g}".format(stat_slope5,
stat_intercept5)
label_line_6 = "y={0:.3g}x+{1:.3g}".format(stat_slope6,
stat_intercept6)
L_labels[0].set_text(label_line_1)
ax = ag_reg_plot.axes[0, 1]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_2)
ax = ag_reg_plot.axes[0, 2]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_3)
ax = ag_reg_plot.axes[1, 0]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_4)
ax = ag_reg_plot.axes[1, 1]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_5)
ax = ag_reg_plot.axes[1, 2]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_6)
ag_reg_plot.set(xlim=(0, 13))
ag_reg_plot.set(ylim=(0.000, 0.003))
# ag_reg_plot.fig.suptitle("RV #%s" % str(replica), y=0.99)
# plt.tight_layout()
ag_reg_plot.savefig(output_dir +
"/ag_lmplot_ADAR_Context_replica%s" % replica,
dpi=300)
plt.close()
data_filter_ag = data_filter_ag[data_filter_ag["Protein"] != "2A"]
data_filter_ag = data_filter_ag[data_filter_ag["Protein"] != "3'UTR"]
data_filter_ag = data_filter_ag[data_filter_ag["Type"] == "Synonymous"]
position_mutation_ag = sns.relplot(
x="Pos",
y="Frequency",
data=data_filter_ag,
hue="5`_ADAR_Preference",
col="passage",
col_wrap=3,
palette=mutation_palette(3, adar=True, ag=True),
hue_order=adar_preference,
height=4,
style="5`_ADAR_Preference",
style_order=["High", "Low", "Intermediate"])
position_mutation_ag.set_axis_labels("", "Variant Frequency")
position_mutation_ag.axes.flat[0].set_yscale('symlog',
linthreshy=10**-4)
position_mutation_ag.axes.flat[0].set_ylim(10**-5, 10**-2)
plt.savefig(output_dir +
"/ag_position_mutation_replica%s.png" % replica,
dpi=300)
plt.close()
columns = ["Mutation", "Replica", "Type", "Slope", "Intercept"]
mutation_rate_ag_df = pd.DataFrame(columns=columns)
mutation_rate_ag_df.loc[0] = [
"High ADAR-like A>G", replica, "Synonymous", stat_slope1,
stat_intercept1
]
mutation_rate_ag_df.loc[1] = [
"Intermediate ADAR-like A>G", replica, "Synonymous", stat_slope2,
stat_intercept2
]
mutation_rate_ag_df.loc[2] = [
"Low ADAR-like A>G", replica, "Synonymous", stat_slope3,
stat_intercept3
]
mutation_rate_ag_df.loc[3] = [
"High ADAR-like A>G", replica, "Non-Synonymous", stat_slope4,
stat_intercept4
]
mutation_rate_ag_df.loc[4] = [
"Intermediate ADAR-like A>G", replica, "Non-Synonymous",
stat_slope5, stat_intercept5
]
mutation_rate_ag_df.loc[5] = [
"Low ADAR-like A>G", replica, "Non-Synonymous", stat_slope6,
stat_intercept6
]
mutation_rate_ag_df.to_csv(output_dir +
"/mutation_rate_ag%s.csv" % replica,
sep=',',
encoding='utf-8')
mutation_rate_ag_df.to_pickle(output_dir +
"/mutation_rate_ag%s.pkl" % replica)
"""U>C Context"""
mutation_uc = sns.catplot("passage",
"frac_and_weight",
data=data_filter_uc,
hue="3`_ADAR_Preference",
palette=mutation_palette(3,
adar=True,
uc=True),
kind="point",
dodge=True,
estimator=weighted_varaint,
orient="v",
col="Type",
join=False,
hue_order=adar_preference,
col_order=type_order_ag)
mutation_uc.set(yscale="log")
mutation_uc.set(ylim=(1 * 10**-5, 1 * 10**-2))
# mutation_uc.set(xticks=["0", "2", "5", "8", "10", "12"])
mutation_uc.set_axis_labels("Passage", "Variant Frequency")
mutation_uc.savefig(
output_dir + "/uc_ADAR_like_Mutation_col_replica%s.png" % replica,
dpi=300)
plt.close()
data_filter_uc_grouped = data_filter_uc.groupby([
"label", "Type", "passage", "replica", "ADAR_grade_three",
"3`_ADAR_Preference"
])["frac_and_weight"].agg(lambda x: weighted_varaint(x))
data_filter_uc_grouped = data_filter_uc_grouped.reset_index()
# print(data_filter_uc_grouped.to_string())
data_filter_uc_grouped = data_filter_uc_grouped.rename(
columns={"frac_and_weight": "Frequency"})
data_filter_uc_grouped["Frequency"] = data_filter_uc_grouped[
"Frequency"].astype(float)
data_filter_uc_grouped = data_filter_uc_grouped[
data_filter_uc_grouped["label"] != "RNA Control\nPrimer ID"]
data_filter_uc_grouped = data_filter_uc_grouped[
data_filter_uc_grouped["label"] != "RNA Control\nRND"]
data_filter_uc_grouped = data_filter_uc_grouped[
data_filter_uc_grouped["replica"] == replica]
data_reg_full_adar_uc = data_filter_uc_grouped[
data_filter_uc_grouped["ADAR_grade_three"] == 1]
data_reg_semi_adar_uc = data_filter_uc_grouped[
data_filter_uc_grouped["ADAR_grade_three"] == 0.5]
data_reg_nonadar_uc = data_filter_uc_grouped[
data_filter_uc_grouped["ADAR_grade_three"] == 0]
data_reg_full_adar_syn_uc = data_reg_full_adar_uc[
data_reg_full_adar_uc["Type"] == "Synonymous"]
data_reg_semi_adar_syn_uc = data_reg_semi_adar_uc[
data_reg_semi_adar_uc["Type"] == "Synonymous"]
data_reg_nonadar_syn_uc = data_reg_nonadar_uc[
data_reg_nonadar_uc["Type"] == "Synonymous"]
data_reg_full_adar_non_syn_uc = data_reg_full_adar_uc[
data_reg_full_adar_uc["Type"] == "Non-Synonymous"]
data_reg_semi_adar_non_syn_uc = data_reg_semi_adar_uc[
data_reg_semi_adar_uc["Type"] == "Non-Synonymous"]
data_reg_nonadar_non_syn_uc = data_reg_nonadar_uc[
data_reg_nonadar_uc["Type"] == "Non-Synonymous"]
stat_slope7, stat_intercept7, r_value7, p_value7, std_err7 = stats.linregress(
data_reg_full_adar_syn_uc['passage'],
data_reg_full_adar_syn_uc['Frequency'])
stat_slope8, stat_intercept8, r_value8, p_value8, std_err8 = stats.linregress(
data_reg_semi_adar_syn_uc['passage'],
data_reg_semi_adar_syn_uc['Frequency'])
stat_slope9, stat_intercept9, r_value9, p_value9, std_err9 = stats.linregress(
data_reg_nonadar_syn_uc['passage'],
data_reg_nonadar_syn_uc['Frequency'])
# data_reg_adar_nonsyn
stat_slope10, stat_intercept10, r_value10, p_value10, std_err10 = stats.linregress(
data_reg_full_adar_non_syn_uc['passage'],
data_reg_full_adar_non_syn_uc['Frequency'])
stat_slope11, stat_intercept11, r_value11, p_value11, std_err11 = stats.linregress(
data_reg_semi_adar_non_syn_uc['passage'],
data_reg_semi_adar_non_syn_uc['Frequency'])
stat_slope12, stat_intercept12, r_value12, p_value12, std_err12 = stats.linregress(
data_reg_nonadar_non_syn_uc['passage'],
data_reg_nonadar_non_syn_uc['Frequency'])
data_filter_uc_grouped = data_filter_uc_grouped.rename(
columns={"passage": "Passage"})
uc_lmplot = sns.lmplot(x="Passage",
y="Frequency",
data=data_filter_uc_grouped,
hue="3`_ADAR_Preference",
markers=["o", "v", "x"],
hue_order=adar_preference,
fit_reg=True,
col="3`_ADAR_Preference",
col_order=adar_preference,
row="Type",
row_order=type_order_ag,
palette=mutation_palette(3, adar=True, uc=True),
line_kws={'label': "Linear Reg"},
legend=True,
height=6)
uc_lmplot.fig.subplots_adjust(wspace=.02)
ax = uc_lmplot.axes[0, 0]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
label_line_7 = "y={0:.3g}x+{1:.3g}".format(stat_slope7,
stat_intercept7)
label_line_8 = "y={0:.3g}x+{1:.3g}".format(stat_slope8,
stat_intercept8)
label_line_9 = "y={0:.3g}x+{1:.3g}".format(stat_slope9,
stat_intercept9)
label_line_10 = "y={0:.3g}x+{1:.3g}".format(stat_slope10,
stat_intercept10)
label_line_11 = "y={0:.3g}x+{1:.3g}".format(stat_slope11,
stat_intercept11)
label_line_12 = "y={0:.3g}x+{1:.3g}".format(stat_slope12,
stat_intercept12)
L_labels[0].set_text(label_line_7)
ax = uc_lmplot.axes[0, 1]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_8)
ax = uc_lmplot.axes[0, 2]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_9)
ax = uc_lmplot.axes[1, 0]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_10)
ax = uc_lmplot.axes[1, 1]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_11)
ax = uc_lmplot.axes[1, 2]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_12)
uc_lmplot.set(xlim=(0, 13))
uc_lmplot.set(ylim=(0.000, 0.003))
# uc_lmplot.fig.suptitle("RV #%s" % str(replica), y=0.99)
# plt.tight_layout()
uc_lmplot.savefig(output_dir +
"/uc_lmplot_ADAR_Context_replica%s" % replica,
dpi=300)
plt.close()
data_filter_uc = data_filter_uc[data_filter_uc["Protein"] != "2A"]
data_filter_uc = data_filter_uc[data_filter_uc["Protein"] != "3'UTR"]
data_filter_uc = data_filter_uc[data_filter_uc["Type"] == "Synonymous"]
position_mutation_uc = sns.relplot(
x="Pos",
y="Frequency",
data=data_filter_uc,
hue="3`_ADAR_Preference",
col="passage",
col_wrap=3,
palette=mutation_palette(3, adar=True, uc=True),
hue_order=adar_preference,
height=4,
style="3`_ADAR_Preference",
style_order=["High", "Low", "Intermediate"])
position_mutation_uc.set_axis_labels("", "Variant Frequency")
position_mutation_uc.axes.flat[0].set_yscale('symlog',
linthreshy=10**-4)
position_mutation_uc.axes.flat[0].set_ylim(10**-5, 10**-2)
plt.savefig(output_dir +
"/uc_position_mutation_replica%s.png" % replica,
dpi=300)
plt.close()
mutation_rate_df1 = pd.read_pickle(output_dir + "/mutation_rate1.pkl")
mutation_rate_df2 = pd.read_pickle(output_dir + "/mutation_rate2.pkl")
mutation_rate_df3 = pd.read_pickle(output_dir + "/mutation_rate3.pkl")
mutation_rate_df_all = pd.concat(
[mutation_rate_df1, mutation_rate_df2, mutation_rate_df3], sort=False)
mutation_rate_df_all.to_csv(output_dir + "/mutation_rate_all.csv",
sep=',',
encoding='utf-8')
mutation_rate_df_all_grouped = mutation_rate_df_all.groupby(
["Mutation", "Type"])["Slope", "Intercept"].agg(np.median)
mutation_rate_df_all_grouped = mutation_rate_df_all_grouped.reset_index()
mutation_rate_df_all_grouped.to_csv(output_dir +
"/mutation_rate_median.csv",
sep=',',
encoding='utf-8')
mutation_rate_ag_df1 = pd.read_pickle(output_dir +
"/mutation_rate_ag1.pkl")
mutation_rate_ag_df2 = pd.read_pickle(output_dir +
"/mutation_rate_ag2.pkl")
mutation_rate_ag_df3 = pd.read_pickle(output_dir +
"/mutation_rate_ag3.pkl")
mutation_rate_ag_df_all = pd.concat(
[mutation_rate_ag_df1, mutation_rate_ag_df2, mutation_rate_ag_df3],
sort=False)
mutation_rate_ag_df_all.to_csv(output_dir + "/mutation_rate_ag_all.csv",
sep=',',
encoding='utf-8')
mutation_rate_ag_df_all.to_pickle(output_dir + "/mutation_rate_ag_all.pkl")
# mutation_rate_ag_df = pd.read_csv(output_dir + "/mutation_rate_ag_all.csv", sep=',', encoding='utf-8')
mutation_rate_ag_df_all_grouped = mutation_rate_ag_df_all.groupby(
["Mutation", "Type"])["Slope", "Intercept"].agg(np.median)
mutation_rate_ag_df_all_grouped = mutation_rate_ag_df_all_grouped.reset_index(
)
mutation_rate_ag_df_all_grouped.to_csv(output_dir +
"/mutation_rate_ag_median.csv",
sep=',',
encoding='utf-8')
def main():
# input_dir = "/Users/odedkushnir/Projects/fitness/AccuNGS/190627_RV_CV/RVB14/"
# input_dir = "/Volumes/STERNADILABHOME$/volume3/okushnir/AccuNGS/20201008RV-202329127/merged/patients/"
input_dir = "/Users/odedkushnir/PhD_Projects/After_review/AccuNGS/RV/patients/"
prefix = "inosine_predict_context_freq0.01"
date = datetime.today().strftime("%Y%m%d")
output_dir = input_dir + "{0}_{1}".format(date, prefix)
try:
os.mkdir(output_dir)
except OSError:
print("Creation of the directory %s failed" % output_dir)
else:
print("Successfully created the directory %s " % output_dir)
data_filter = pd.read_pickle(input_dir + prefix + "/data_filter.pkl")
data_filter_ag = pd.read_pickle(input_dir + prefix + "/data_filter_ag.pkl")
data_filter_uc = pd.read_pickle(input_dir + prefix + "/data_filter_uc.pkl")
data_filter["label"] = np.where(
data_filter["label"] == "RNA Control\nPrimer ID", "RNA\nControl",
data_filter["label"])
#Plots
label_order = [
"RNA\nControl", "p3 Cell Culture\nControl", "Patient-1", "Patient-4",
"Patient-5", "Patient-9", "Patient-16", "Patient-17", "Patient-20"
]
mutation_order = [
"A>G", "U>C", "G>A", "C>U", "A>C", "U>G", "A>U", "U>A", "G>C", "C>G",
"C>A", "G>U"
]
transition_order = ["A>G", "U>C", "G>A", "C>U"]
type_order1 = ["Synonymous", "Non-Synonymous", "Premature Stop Codon"]
context_order = ["UpA", "ApA", "CpA", "GpA"]
type_order2 = ["Synonymous", "Non-Synonymous"]
context_order_uc = ["UpA", "UpU", "UpG", "UpC"]
type_order_ag = ["Synonymous", "Non-Synonymous", "NonCodingRegion"]
adar_preference = ["High", "Intermediate", "Low"]
plus_minus = u"\u00B1"
pairs = [(("RNA\nControl", "A>G"), ("RNA\nControl", "G>A")),
(("p3 Cell Culture\nControl", "A>G"), ("p3 Cell Culture\nControl",
"G>A")),
(("Patient-1", "A>G"), ("Patient-1", "G>A")),
(("Patient-4", "A>G"), ("Patient-4", "G>A")),
(("Patient-5", "A>G"), ("Patient-5", "G>A")),
(("Patient-9", "A>G"), ("Patient-9", "G>A")),
(("Patient-16", "A>G"), ("Patient-16", "G>A")),
(("Patient-17", "A>G"), ("Patient-17", "G>A")),
(("Patient-20", "A>G"), ("Patient-20", "G>A")),
(("RNA\nControl", "A>G"), ("RNA\nControl", "U>C")),
(("p3 Cell Culture\nControl", "A>G"), ("p3 Cell Culture\nControl",
"U>C")),
(("Patient-1", "A>G"), ("Patient-1", "U>C")),
(("Patient-4", "A>G"), ("Patient-4", "U>C")),
(("Patient-5", "A>G"), ("Patient-5", "U>C")),
(("Patient-9", "A>G"), ("Patient-9", "U>C")),
(("Patient-16", "A>G"), ("Patient-16", "U>C")),
(("Patient-17", "A>G"), ("Patient-17", "U>C")),
(("Patient-20", "A>G"), ("Patient-20", "U>C")),
(("RNA\nControl", "A>G"), ("RNA\nControl", "C>U")),
(("p3 Cell Culture\nControl", "A>G"), ("p3 Cell Culture\nControl",
"C>U")),
(("Patient-1", "A>G"), ("Patient-1", "C>U")),
(("Patient-4", "A>G"), ("Patient-4", "C>U")),
(("Patient-5", "A>G"), ("Patient-5", "C>U")),
(("Patient-9", "A>G"), ("Patient-9", "C>U")),
(("Patient-16", "A>G"), ("Patient-16", "C>U")),
(("Patient-17", "A>G"), ("Patient-17", "C>U")),
(("Patient-20", "A>G"), ("Patient-20", "C>U"))]
# g1 = sns.catplot(x="label", y="frac_and_weight", data=data_filter, hue="Mutation", order=label_order, palette="tab20",
# kind="point", dodge=True, hue_order=mutation_order, join=False, estimator=weighted_varaint,
# orient="v")
# g1.set_axis_labels("", "Variant Frequency {} CI=95%".format(plus_minus))
# g1.set_xticklabels(fontsize=9, rotation=90)
# g1.set(yscale='log')
# # g1.set(ylim=(10**-7, 10**-3))
#
# # plt.show()
# g1.savefig(output_dir + "/All_Mutations_point_plot", dpi=300)
# plt.close()
g2 = sns.catplot(x="label",
y="frac_and_weight",
data=data_filter,
hue="Mutation",
order=label_order,
palette=mutation_palette(4),
kind="point",
dodge=0.5,
hue_order=transition_order,
join=False,
estimator=weighted_varaint,
orient="v",
legend=True)
g2.set_axis_labels("", "Variant Frequency {} CI=95%".format(plus_minus))
g2.set(yscale='log')
g2.set(ylim=(10**-5, 10**-3))
# g2.set_yticklabels(fontsize=12)
g2.set_xticklabels(fontsize=10, rotation=90)
# plt.show()
# g2.savefig("/Users/odedkushnir/Google Drive/Studies/PhD/MyPosters/20190924 GGE/plots/Transition_Mutations_point_plot_RV", dpi=300)
g2.savefig(output_dir + "/Transition_Mutations_point_plot", dpi=300)
# g2.savefig("/Users/odedkushnir/Google Drive/Studies/PhD/Prgress reports/20200913 Final report/plots" +
# "/Fig9a_Transition_Mutations_point_plot_Patients", dpi=300)
plt.close()
data_filter["label"] = data_filter["label"].astype(str)
data_filter["Frequency"] = data_filter["Frequency"].astype(float)
passage_g = sns.boxplot(x="label",
y="Frequency",
data=data_filter,
hue="Mutation",
order=label_order,
palette=mutation_palette(4),
dodge=True,
hue_order=transition_order)
passage_g.set_yscale('log')
passage_g.set_ylim(10**-6, 10**-1)
passage_g.set(xlabel="", ylabel="Variant Frequency")
passage_g.set_xticklabels(labels=label_order, fontsize=10, rotation=90)
annot = Annotator(passage_g,
pairs,
x="label",
y="Frequency",
hue="Mutation",
data=data_filter,
order=label_order,
hue_order=transition_order)
annot.configure(test='t-test_welch',
text_format='star',
loc='outside',
verbose=2,
comparisons_correction="Bonferroni")
annot.apply_test()
file_path = output_dir + "/sts.csv"
with open(file_path, "w") as o:
with contextlib.redirect_stdout(o):
passage_g, test_results = annot.annotate()
plt.legend(bbox_to_anchor=(1.05, 0.5), loc=2, borderaxespad=0.)
plt.tight_layout()
plt.savefig(output_dir + "/Transition_Mutations_box_stat_plot_patients",
dpi=300)
plt.close()
# g_rna = sns.catplot(x="RNA", y="frac_and_weight", data=data_filter, hue="Mutation", order=rna_order,
# palette="tab20", kind="point", dodge=True, hue_order=transition_order, join=False, estimator=weighted_varaint,
# orient="v", legend=True)
# g_rna.set_axis_labels("", "Variant Frequency")
# g_rna.set(yscale='log')
# g_rna.set(ylim=(10 ** -6, 10 ** -2))
# # g2.set_yticklabels(fontsize=12)
# g_rna.set_xticklabels(fontsize=10, rotation=45)
# plt.show()
# g2.savefig("/Users/odedkushnir/Google Drive/Studies/PhD/MyPosters/20190924 GGE/plots/Transition_Mutations_point_plot_RV", dpi=300)
# g_rna.savefig(output_dir + "/Transition_Mutations_point_RNA_plot", dpi=300)
# plt.close()
# A>G Prev Context
flatui = ["#3498db", "#9b59b6"]
g5 = sns.catplot("label",
"frac_and_weight",
data=data_filter_ag,
hue="ADAR_like",
order=label_order,
palette=mutation_palette(2),
kind="point",
dodge=True,
hue_order=[True, False],
estimator=weighted_varaint,
orient="v",
col="Type",
join=False,
col_order=type_order2)
g5.set_axis_labels("", "Variant Frequency {} CI=95%".format(plus_minus))
g5.set(yscale='log')
g5.set(ylim=(7 * 10**-7, 4 * 10**-3))
g5.set_xticklabels(rotation=90)
# plt.show()
g5.savefig(output_dir + "/Context_point_plot", dpi=300)
# g5.savefig("/Users/odedkushnir/Google Drive/Studies/PhD/Prgress reports/20200913 Final report/plots" +
# "/Fig9b_Context_point_plot_Patients", dpi=300)
plt.close()
mutation_ag = sns.catplot("label",
"frac_and_weight",
data=data_filter_ag,
hue="5`_ADAR_Preference",
palette=mutation_palette(3, adar=True, ag=True),
kind="point",
dodge=True,
estimator=weighted_varaint,
order=label_order,
orient="v",
col="Type",
join=False,
col_order=type_order_ag,
hue_order=adar_preference)
mutation_ag.set(yscale="log")
mutation_ag.set(ylim=(1 * 10**-5, 1 * 10**-2))
mutation_ag.set_xticklabels(rotation=90)
mutation_ag.fig.suptitle("A>G ADAR_like Mutation in RV patients", y=0.99)
plt.subplots_adjust(top=0.85)
mutation_ag.set_axis_labels(
"", "Variant Frequency {} CI=95%".format(plus_minus))
mutation_ag.savefig(output_dir + "/ag_ADAR_like_Mutation_col_patients.png",
dpi=300)
plt.close()
g6 = sns.catplot("label",
"frac_and_weight",
data=data_filter_ag,
hue="ADAR_like",
order=label_order,
palette=mutation_palette(2),
kind="point",
dodge=True,
hue_order=[True, False],
estimator=weighted_varaint,
orient="v",
join=False)
g6.set_axis_labels("", "Variant Frequency {} CI=95%".format(plus_minus))
g6.set(yscale='log')
g6.set(ylim=(7 * 10**-7, 4 * 10**-3))
g6.set_xticklabels(rotation=90)
# plt.show()
g6.savefig(output_dir + "/Context_point_all_mutations_type_plot", dpi=300)
plt.close()
g9 = sns.catplot("label",
"frac_and_weight",
data=data_filter_uc,
hue="Next",
order=label_order,
palette="tab20",
hue_order=context_order_uc,
estimator=weighted_varaint,
orient="v",
dodge=True,
kind="point",
col="Type",
join=False,
col_order=type_order2)
g9.set_axis_labels("", "Variant Frequency {} CI=95%".format(plus_minus))
g9.set(yscale='log')
g9.set(ylim=(10**-5, 10**-2))
g9.set_xticklabels(rotation=90)
# plt.show()
g9.savefig(output_dir + "/UC_Context_point_plot", dpi=300)
plt.close()
data_filter_ag_grouped = data_filter_ag.groupby(
["ADAR_like", "label",
"Type"])["frac_and_weight"].agg(lambda x: weighted_varaint(x))
data_filter_ag_grouped = data_filter_ag_grouped.reset_index()
data_filter_ag_grouped = data_filter_ag_grouped.rename(
columns={"frac_and_weight": "Frequency"})
data_filter_ag_grouped["Frequency"] = data_filter_ag_grouped[
"Frequency"].astype(float)
print(data_filter_ag_grouped.to_string())
data_filter_ag_grouped_silent = data_filter_ag_grouped[
data_filter_ag_grouped["Type"] == "Synonymous"]
data_filter_ag_grouped_silent = data_filter_ag_grouped_silent[
data_filter_ag_grouped_silent["label"] == "Cell Cultureֿ\nControl"]
def linear_reg(data_filter, output_dir, transition_order, type_order, virus, replica, cu=True, ag=True, uc=True,
ga= True, output_file="/mutation_rate"):
"""
:param data_filter:
:param output_dir:
:param transition_order:
:param type_order:
:param virus:
:param replica:
:param cu:
:param output_file:
:return:
"""
# data_filter = data_filter.groupby(["label", "passage", "Type", "Mutation", "replica"])[
# "frac_and_weight"].agg(
# lambda x: weighted_varaint(x))
# data_filter = data_filter.reset_index()
# data_filter = data_filter.rename(columns={"frac_and_weight": "Frequency"})
# data_filter["Frequency"] = data_filter["Frequency"].astype(float)
data_filter = data_filter[data_filter["label"] != "RNA Control\nPrimer ID"]
data_filter = data_filter[data_filter["label"] != "RNA Control\nRND"]
data_filter = data_filter[data_filter["replica"] == replica]
# print(data_filter.to_string())
data_reg_ag = data_filter[data_filter["Mutation"] == "A>G"]
data_reg_uc = data_filter[data_filter["Mutation"] == "U>C"]
data_reg_ga = data_filter[data_filter["Mutation"] == "G>A"]
data_reg_cu = data_filter[data_filter["Mutation"] == "C>U"]
data_reg_ag_syn = data_reg_ag[data_reg_ag["Type"] == "Synonymous"]
data_reg_uc_syn = data_reg_uc[data_reg_uc["Type"] == "Synonymous"]
data_reg_ga_syn = data_reg_ga[data_reg_ga["Type"] == "Synonymous"]
data_reg_cu_syn = data_reg_cu[data_reg_cu["Type"] == "Synonymous"]
data_reg_ag_non_syn = data_reg_ag[data_reg_ag["Type"] == "Non-Synonymous"]
data_reg_uc_non_syn = data_reg_uc[data_reg_uc["Type"] == "Non-Synonymous"]
data_reg_ga_non_syn = data_reg_ga[data_reg_ga["Type"] == "Non-Synonymous"]
data_reg_cu_non_syn = data_reg_cu[data_reg_cu["Type"] == "Non-Synonymous"]
data_reg_ga_pmsc = data_reg_ga[data_reg_ga["Type"] == "Premature Stop Codon"]
data_reg_cu_pmsc = data_reg_cu[data_reg_cu["Type"] == "Premature Stop Codon"]
if ag==True:
slope1, intercept1, r_value1, p_value1, std_err1 = stats.linregress(data_reg_ag_syn['passage'],
data_reg_ag_syn
['Frequency'])
slope5, intercept5, r_value5, p_value5, std_err5 = stats.linregress(data_reg_ag_non_syn['passage'],
data_reg_ag_non_syn[
'Frequency'])
if uc==True:
slope2, intercept2, r_value2, p_value2, std_err2 = stats.linregress(data_reg_uc_syn['passage'],
data_reg_uc_syn[
'Frequency'])
slope6, intercept6, r_value6, p_value6, std_err6 = stats.linregress(data_reg_uc_non_syn['passage'],
data_reg_uc_non_syn[
'Frequency'])
if ga == True:
slope3, intercept3, r_value3, p_value3, std_err3 = stats.linregress(data_reg_ga_syn['passage'],
data_reg_ga_syn[
'Frequency'])
slope7, intercept7, r_value7, p_value7, std_err7 = stats.linregress(data_reg_ga_non_syn['passage'],
data_reg_ga_non_syn[
'Frequency'])
slope11, intercept11, r_value11, p_value11, std_err11 = stats.linregress(data_reg_ga_pmsc['passage'],
data_reg_ga_pmsc[
'Frequency'])
if cu == True:
slope4, intercept4, r_value4, p_value4, std_err4 = stats.linregress(data_reg_cu_syn['passage'],
data_reg_cu_syn[
'Frequency'])
slope8, intercept8, r_value8, p_value8, std_err8 = stats.linregress(data_reg_cu_non_syn['passage'],
data_reg_cu_non_syn[
'Frequency'])
slope12, intercept12, r_value12, p_value12, std_err12 = stats.linregress(data_reg_cu_pmsc['passage'],
data_reg_cu_pmsc[
'Frequency'])
data_filter = data_filter.rename(columns={"passage": "Passage"})
reg_plot = sns.lmplot(x="Passage", y="Frequency", data=data_filter, hue="Mutation",
hue_order=transition_order, fit_reg=True, col="Mutation",
col_order=transition_order, row="Type", row_order=type_order, palette=mutation_palette(4),
line_kws={'label': "Linear Reg"}, legend=True, height=6) # markers=["o", "v", "x"]
if ag == None:
slope1, intercept1, r_value1, p_value1, std_err1 = 0, 0, 0, 0, 0
slope5, intercept5, r_value5, p_value5, std_err5 = 0, 0, 0, 0, 0
if uc == None:
slope2, intercept2, r_value2, p_value2, std_err2 = 0, 0, 0, 0, 0
slope6, intercept6, r_value6, p_value6, std_err6 = 0, 0, 0, 0, 0
if ga == None:
slope3, intercept3, r_value3, p_value3, std_err3 = 0, 0, 0, 0, 0
slope7, intercept7, r_value7, p_value7, std_err7 = 0, 0, 0, 0, 0
slope11, intercept11, r_value11, p_value11, std_err11 = 0, 0, 0, 0, 0
if cu == None:
slope4, intercept4, r_value4, p_value4, std_err4 = 0, 0, 0, 0, 0
slope8, intercept8, r_value8, p_value8, std_err8 = 0, 0, 0, 0, 0
slope12, intercept12, r_value12, p_value12, std_err12 = 0, 0, 0, 0, 0
label_line_1 = "y={0:.3g}x+{1:.3g} pval={2:.3g}".format(slope1, intercept1, p_value1)
label_line_2 = "y={0:.3g}x+{1:.3g} pval={2:.3g}".format(slope2, intercept2, p_value2)
label_line_3 = "y={0:.3g}x+{1:.3g} pval={2:.3g}".format(slope3, intercept3, p_value3)
label_line_4 = "y={0:.3g}x+{1:.3g} pval={2:.3g}".format(slope4, intercept4, p_value4)
label_line_5 = "y={0:.3g}x+{1:.3g} pval={2:.3g}".format(slope5, intercept5, p_value5)
label_line_6 = "y={0:.3g}x+{1:.3g} pval={2:.3g}".format(slope6, intercept6, p_value6)
label_line_7 = "y={0:.3g}x+{1:.3g} pval={2:.3g}".format(slope7, intercept7, p_value7)
label_line_8 = "y={0:.3g}x+{1:.3g} pval={2:.3g}".format(slope8, intercept8, p_value8)
label_line_11 = "y={0:.3g}x+{1:.3g} pval={2:.3g}".format(slope11, intercept11, p_value11)
label_line_12 = "y={0:.3g}x+{1:.3g} pval={2:.3g}".format(slope12, intercept12, p_value12)
reg_plot.fig.subplots_adjust(wspace=.02)
if ag == True:
ax = reg_plot.axes[0, 0]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_1)
if uc == True:
ax = reg_plot.axes[0, 1]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_2)
if ga == True:
ax = reg_plot.axes[0, 2]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_3)
if cu == True:
ax = reg_plot.axes[0, 3]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_4)
if ag == True:
ax = reg_plot.axes[1, 0]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_5)
if uc == True:
ax = reg_plot.axes[1, 1]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_6)
if ga == True:
ax = reg_plot.axes[1, 2]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_7)
if cu == True:
ax = reg_plot.axes[1, 3]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_8)
if ga == True:
ax = reg_plot.axes[2, 2]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_11)
if cu == True:
ax = reg_plot.axes[2, 3]
ax.legend()
leg = ax.get_legend()
leg._loc = 2
L_labels = leg.get_texts()
L_labels[0].set_text(label_line_12)
if virus == "RVB14":
reg_plot.set(xlim=(0, 13))
reg_plot.set(ylim=(0.000, 0.01))
# reg_plot.fig.suptitle("RV #%s" % str(replica), y=0.99)
plt.tight_layout()
reg_plot.savefig(output_dir + output_file + "_lmplot_%s.png" % replica, dpi=300)
plt.close()
columns = ["Mutation", "Type", "Slope", "Intercept", "p-val"]
mutation_rate_df = pd.DataFrame(columns=columns)
mutation_rate_df.loc[0] = ["A>G", "Synonymous", slope1, intercept1, p_value1]
mutation_rate_df.loc[1] = ["U>C", "Synonymous", slope2, intercept2, p_value2]
mutation_rate_df.loc[2] = ["G>A", "Synonymous", slope3, intercept3, p_value3]
mutation_rate_df.loc[3] = ["C>U", "Synonymous", slope4, intercept4, p_value3]
mutation_rate_df.loc[4] = ["A>G", "Non-Synonymous", slope5, intercept5, p_value5]
mutation_rate_df.loc[5] = ["U>C", "Non-Synonymous", slope6, intercept6, p_value6]
mutation_rate_df.loc[6] = ["G>A", "Non-Synonymous", slope7, intercept7, p_value7]
mutation_rate_df.loc[7] = ["C>U", "Non-Synonymous", slope8, intercept8, p_value8]
mutation_rate_df.loc[8] = ["G>A", "Pre Mature Stop Codon", slope11, intercept11, p_value11]
mutation_rate_df.loc[9] = ["C>U", "Pre Mature Stop Codon", slope12, intercept12, p_value11]
mutation_rate_df["Virus"] = virus
mutation_rate_df["Replica"] = replica
mutation_rate_df.to_csv(output_dir + output_file + ".csv", sep=',', encoding='utf-8')
mutation_rate_df.to_pickle(output_dir + output_file + str(replica) + ".pkl")
return data_filter
def plots_for_srr(input_dir, output_dir, virus_path, virus):
data_mutations = pd.read_csv(input_dir + virus_path + "data_mutation.csv")
# data_mutations["Mutation"] = data_mutations["Mutation"].apply(lambda x: x.split("->")[0] + ">" + x.split("->")[1])
mutation_order = ["A>G", "U>C", "C>U", "G>A", "A>U", "A>C", "U>A", "U>G", "C>A", "C>G", "G>U", "G>C"]
transition_order = ["A>G", "U>C", "C>U", "G>A"]
type_order = ["Synonymous", "Non-Synonymous", "Premature Stop Codon"]
# data_mutations = data_mutations[data_mutations["pval"] < 0.01]
data_mutations = data_mutations[data_mutations["Prob"] > 0]
# plt.style.use('classic')
sns.set_style("ticks")
mypalette = ["#3498db", "#9b59b6"]
adar_order = ("ADAR-like", "Non\nADAR-like")
# All Mutations
with sns.plotting_context(rc={"legend.fontsize": 6}):
all_plot = sns.catplot(x="Mutation", y="Frequency", hue="label", data=data_mutations, palette="tab20",
order=mutation_order, flierprops={"marker": "."},
kind="box", col="Type", col_order=type_order)
all_plot.set(yscale='log')
all_plot._legend.set_title('')
all_plot.set(xlabel="")
all_plot.fig.suptitle("All Mutations variant frequencies in %s" % virus, y=0.99)
plt.savefig(output_dir + "All_Mutations_variant_frequencies_in_%s.png" % virus, dpi=300)
plt.close()
# Transitions mutations plot
with sns.plotting_context(rc={"legend.fontsize": 6}):
transiotion_plot = sns.catplot(x="Mutation", y="Frequency", hue="label", data=data_mutations, palette="tab20",
order=transition_order, flierprops={"marker": "."},
kind="box")
transiotion_plot.set(yscale='log')
transiotion_plot._legend.set_title('')
transiotion_plot.set(xlabel="")
transiotion_plot.fig.suptitle("Transitions variant frequency in %s" % virus, y=0.99)
plt.savefig(output_dir + "Transitions_variant_frequency_in_%s.png" % virus, dpi=300)
plt.close()
##Context
#5'
data_context = pd.read_csv(input_dir + virus_path + "data_XpA_by_mutation.csv")
data_context["no_variants"] = data_context["Freq"] * data_context["Read_count"]
data_context["freq_and_weight"] = list(zip(data_context.no_variants, data_context.Read_count))
if virus == "EnteroA":
data_context = data_context.loc[data_context.Organism != "Coxsackievirus A16"]
# data_context["Mutation"] = data_context["Mutation"].apply(lambda x: x.split("->")[0] + ">" + x.split("->")[1])
data_context = data_context.rename(columns={"Freq": "Frequency"})
data_context['Prev'].replace('AA', 'ApA', inplace=True)
data_context['Prev'].replace('UA', 'UpA', inplace=True)
data_context['Prev'].replace('CA', 'CpA', inplace=True)
data_context['Prev'].replace('GA', 'GpA', inplace=True)
data_context['Type'].replace('Synonymous', 'Silent', inplace=True)
data_context['Type'].replace('Non-Synonymous', 'Missense', inplace=True)
# data_context_rv = data_context_rv[data_context_rv["pval"] < 0.01]
data_context = data_context[data_context["Prob"] > 0]
context_order = ["UpA", "ApA", "CpA", "GpA"]
data_adar = data_context.loc[data_context.Type == "Silent"]
# print(type(data_adar["Context"]))
data_adar["ADAR_like"] = data_adar.Prev.str.contains('UpA') | data_adar.Prev.str.contains('ApA')
data_adar["ADAR_like"] = np.where(data_adar["ADAR_like"] == True, "ADAR-like", "Non\nADAR-like")
no_organism = data_adar.Organism.value_counts()
#3'
data_context_3 = pd.read_csv(input_dir + virus_path + "data_UpX_by_mutation.csv")
if virus == "EnteroA":
data_context = data_context.loc[data_context.Organism != "Coxsackievirus A16"]
# data_context["Mutation"] = data_context["Mutation"].apply(lambda x: x.split("->")[0] + ">" + x.split("->")[1])
data_context_3 = data_context_3.rename(columns={"Freq": "Frequency"})
data_context_3['Next'].replace('UA', 'UpA', inplace=True)
data_context_3['Next'].replace('UU', 'UpU', inplace=True)
data_context_3['Next'].replace('UC', 'UpC', inplace=True)
data_context_3['Next'].replace('UG', 'UpG', inplace=True)
data_context_3['Type'].replace('Synonymous', 'Silent', inplace=True)
data_context_3['Type'].replace('Non-Synonymous', 'Missense', inplace=True)
# data_context_rv = data_context_rv[data_context_rv["pval"] < 0.01]
data_context_3 = data_context_3[data_context_3["Prob"] > 0]
context_order_3 = ["UpA", "UpU", "UpC", "UpG"]
data_adar_3 = data_context_3.loc[data_context_3.Type == "Silent"]
# print(type(data_adar["Context"]))
data_adar_3["ADAR_like"] = data_adar_3.Next.str.contains('UpA') | data_adar_3.Next.str.contains('UpU')
data_adar_3["ADAR_like"] = np.where(data_adar_3["ADAR_like"] == True, "ADAR-like", "Non\nADAR-like")
no_organism = data_adar.Organism.value_counts()
# 5’ neighbors preferences
#stat
data_adar["log10_Frequency"] = data_adar["Frequency"].apply(lambda x: np.log10(x))
data_adar["log10_Frequency"] = data_adar["log10_Frequency"].astype(float)
if virus == "EnteroA":
context_stat_plot = sns.catplot("ADAR_like", "Frequency", data=data_adar, palette=mutation_palette(2), col="Organism",
kind="boxen", col_wrap=int(len(no_organism)/2), order=adar_order) #flierprops={"marker": "."},
else:
context_stat_plot = sns.catplot("ADAR_like", "Frequency", data=data_adar, palette=mutation_palette(2), col="Organism",
kind="boxen", col_wrap=len(no_organism), order=adar_order) #flierprops={"marker": "."},
context_stat_plot.set(yscale='log')
context_stat_plot.set(ylim=(10 ** -4, 10 ** -2))
# context_stat_plot.axes.flat[0].set_yscale('symlog', linthreshy=10 ** -5)
context_stat_plot.set_axis_labels("", "Variant Frequency")
organism_list = data_adar["Organism"].unique()
for ax in context_stat_plot.axes.flat:
print(ax.get_title())
for organism in organism_list:
if ax.get_title().split(" = ")[-1] == organism:
ax.set_title("%s" % organism, pad=20)
old_statannot.add_stat_annotation(ax, data=data_adar[data_adar["Organism"] == organism], x="ADAR_like", y="Frequency",
boxPairList=[("ADAR-like", "Non\nADAR-like")], test='Mann-Whitney', textFormat='star', loc='inside',
verbose=0, lineOffsetToBox=2, lineHeight=0, stack=False, useFixedOffset=True)
# context_stat_plot.legend(bbox_to_anchor=(1.05, 0.5), loc=3, borderaxespad=0., fontsize='small')
plt.tight_layout()
plt.savefig(output_dir + "Boxenplot_Stat_synonymous_variant_frequency_in_%s.png" % virus, dpi=300)
plt.close()
if virus == "EnteroA":
context_stat_plot = sns.catplot(x="ADAR_like", y="freq_and_weight", data=data_adar, palette=mutation_palette(2),
col="Organism", kind="point", dodge=True, col_wrap=int(len(no_organism)/2),
order=adar_order, join=False, estimator=weighted_varaint, orient="v",
legend=True) #flierprops={"marker": "."},
else:
context_stat_plot = sns.catplot(x="ADAR_like", y="freq_and_weight", data=data_adar, palette=mutation_palette(2),
col="Organism", kind="point", col_wrap=len(no_organism), order=adar_order,
join=False, estimator=weighted_varaint, orient="v", legend=True) #flierprops={"marker": "."}
context_stat_plot.set(yscale='log')
context_stat_plot.set(ylim=(10 ** -4, 10 ** -2))
# context_stat_plot.axes.flat[0].set_yscale('symlog', linthreshy=10 ** -5)
context_stat_plot.set_axis_labels("", "Variant Frequency")
organism_list = data_adar["Organism"].unique()
# context_stat_plot.legend(bbox_to_anchor=(1.05, 0.5), loc=3, borderaxespad=0., fontsize='small')
plt.tight_layout()
plt.savefig(output_dir + "Pointplot_Stat_synonymous_variant_frequency_in_%s.png" % virus, dpi=300)
plt.close()
# 3’ neighbors preferences
#stat
if virus == "EnteroA":
context_stat_plot = sns.catplot("ADAR_like", "Frequency", data=data_adar_3, palette=mutation_palette(2), col="Organism",
flierprops={"marker": "."}, kind="box", col_wrap=int(len(no_organism)/2), order=adar_order)
else:
context_stat_plot = sns.catplot("ADAR_like", "Frequency", data=data_adar_3, palette=mutation_palette(2), col="Organism",
flierprops={"marker": "."}, kind="box", col_wrap=len(no_organism), order=adar_order)
context_stat_plot.set(yscale='log')
context_stat_plot.set(ylim=(10 ** -5, 0))
context_stat_plot.set_axis_labels("", "Variant Frequency")
organism_list = data_adar["Organism"].unique()
for ax in context_stat_plot.axes.flat:
print(ax.get_title())
for organism in organism_list:
if ax.get_title().split(" = ")[-1] == organism:
ax.set_title("3' ADAR context in %s" % organism, pad=20)
old_statannot.add_stat_annotation(ax, data=data_adar_3[data_adar_3["Organism"] == organism], x="ADAR_like", y="Frequency",
boxPairList=[("ADAR-like", "Non\nADAR-like")], test='Mann-Whitney', textFormat='star', loc='inside',
verbose=0, lineOffsetToBox=2, lineHeight=0, stack=False, useFixedOffset=True)
# context_stat_plot.legend(bbox_to_anchor=(1.05, 0.5), loc=3, borderaxespad=0., fontsize='small')
plt.tight_layout()
plt.savefig(output_dir + "Stat_synonymous_variant_frequency_in_%s_3.png" % virus, dpi=300)
plt.close()
def main():
# input_dir = "/Users/odedkushnir/Projects/fitness/AccuNGS/190627_RV_CV/RVB14/"
# input_dir = "/Volumes/STERNADILABHOME$/volume3/okushnir/AccuNGS/20201008RV-202329127/merged/capsid/"
input_dir = "/Users/odedkushnir/PhD_Projects/After_review/AccuNGS/RV/capsid/"
prefix = "inosine_predict_context"
date = datetime.today().strftime("%Y%m%d")
output_dir = input_dir + "{0}_{1}".format(date, prefix)
try:
os.mkdir(output_dir)
except OSError:
print("Creation of the directory %s failed" % output_dir)
else:
print("Successfully created the directory %s " % output_dir)
data_filter = pd.read_pickle(input_dir + prefix + "/data_filter.pkl")
data_filter_ag = pd.read_pickle(input_dir + prefix + "/data_filter_ag.pkl")
data_filter_uc = pd.read_pickle(input_dir + prefix + "/data_filter_uc.pkl")
data_filter_replica1 = data_filter[
(data_filter["label"] == "Capsid-31-Amicon") |
(data_filter["label"] == "Free-31-Amicon") |
(data_filter["label"] == "RNA Control\nPrimer ID") |
(data_filter["label"] == "p8 Mixed Population")]
data_filter_replica1["RNA"] = np.where(
(data_filter_replica1["RNA"] == "Capsid"), "p9 Capsid #1",
data_filter_replica1["RNA"])
data_filter_replica1["RNA"] = np.where(
(data_filter_replica1["RNA"] == "Free"), "p9 Free #1",
data_filter_replica1["RNA"])
data_filter_replica1["RNA"] = np.where(
(data_filter_replica1["RNA"] == "p8 Mixed Population"),
"p8 Mixed\nPopulation", data_filter_replica1["RNA"])
data_filter_replica1.to_csv(input_dir + prefix + "/data_filter_rep1.csv",
sep=",",
encoding='utf-8')
data_filter_ag_replica1 = data_filter_ag[
(data_filter_ag["label"] == "Capsid-31-Amicon") |
(data_filter_ag["label"] == "Free-31-Amicon") |
(data_filter_ag["label"] == "RNA Control\nPrimer ID") |
(data_filter_ag["label"] == "p8 Mixed Population")]
data_filter_ag_replica1["RNA"] = np.where(
(data_filter_ag_replica1["RNA"] == "Capsid"), "p9 Capsid #1",
data_filter_ag_replica1["RNA"])
data_filter_ag_replica1["RNA"] = np.where(
(data_filter_ag_replica1["RNA"] == "Free"), "p9 Free #1",
data_filter_ag_replica1["RNA"])
data_filter_ag_replica1["RNA"] = np.where(
(data_filter_ag_replica1["RNA"] == "p8 Mixed Population"),
"p8 Mixed\nPopulation", data_filter_ag_replica1["RNA"])
data_filter_ag_replica1.to_csv(input_dir + prefix +
"/data_filter_ag_rep1.csv",
sep=",",
encoding='utf-8')
#Plots
label_order = [
"RNA Control\nPrimer ID", "p8 Mixed Population", "Capsid-31-Amicon",
"Capsid-32-Ultra", "Capsid-33-Ultra", "Free-31-Amicon",
"Free-32-Ultra", "Free-33-Ultra"
] #
rna_order_replica1 = [
"RNA Control\nPrimer ID", "p8 Mixed\nPopulation", "p9 Capsid #1",
"p9 Free #1"
]
transition_order = ["A>G", "U>C", "G>A", "C>U"]
type_order_ag = ["Synonymous", "Non-Synonymous"]
adar_preference = ["High", "Intermediate", "Low"]
rna_order = [
"RNA Control\nPrimer ID", "p8 Mixed Population", "Capsid", "Free"
]
type_order = ["Synonymous", "Non-Synonymous", "Premature Stop Codon"]
mutation_order = [
"A>G", "U>C", "G>A", "C>U", "A>C", "U>G", "A>U", "U>A", "G>C", "C>G",
"C>A", "G>U"
]
context_order_uc = ["UpU", "UpA", "UpC", "UpG"]
context_order = ["UpA", "ApA", "CpA", "GpA"]
type_order = ["Synonymous", "Non-Synonymous"]
plus_minus = u"\u00B1"
# g1 = sns.catplot(x="label", y="frac_and_weight", data=data_filter, hue="Mutation", order=label_order, palette="tab20",
# kind="point", dodge=False, hue_order=mutation_order, join=True, estimator=weighted_varaint,
# orient="v")
# g1.set_axis_labels("", "Variant Frequency")
# g1.set_xticklabels(fontsize=9, rotation=45)
# g1.set(yscale='log')
# g1.set(ylim=(10**-7, 10**-3))
# g1.savefig(output_dir + "/All_Mutations_point_plot", dpi=300)
# plt.close()
g2 = sns.catplot(x="label",
y="frac_and_weight",
data=data_filter,
hue="Mutation",
order=label_order,
palette=mutation_palette(4),
kind="point",
dodge=True,
hue_order=transition_order,
join=False,
estimator=weighted_varaint,
orient="v",
legend=True)
g2.set_axis_labels("", "Variant Frequency {} CI=95%".format(plus_minus))
g2.set(yscale='log')
g2.set(ylim=(10**-6, 10**-2))
# g2.set_yticklabels(fontsize=12)
g2.set_xticklabels(fontsize=10, rotation=90)
# plt.show()
# g2.savefig("/Users/odedkushnir/Google Drive/Studies/PhD/MyPosters/20190924 GGE/plots/Transition_Mutations_point_plot_RV", dpi=300)
g2.savefig(output_dir + "/Transition_Mutations_label_point_plot", dpi=300)
plt.close()
g_rna = sns.catplot(x="RNA",
y="frac_and_weight",
data=data_filter_replica1,
hue="Mutation",
order=rna_order_replica1,
palette=mutation_palette(4),
kind="point",
dodge=True,
hue_order=transition_order,
join=False,
estimator=weighted_varaint,
orient="v",
legend=True)
g_rna.set_axis_labels("", "Variant Frequency {} CI=95%".format(plus_minus))
g_rna.set(yscale='log')
g_rna.set(ylim=(10**-6, 10**-2))
g_rna.savefig(output_dir + "/Transition_Mutations_RNA_point_plot", dpi=300)
plt.close()
# A>G Prev Context
g4 = sns.catplot("label",
"frac_and_weight",
data=data_filter_ag,
hue="5`_ADAR_Preference",
order=label_order,
palette=mutation_palette(3, adar=True, ag=True),
kind="point",
dodge=True,
hue_order=adar_preference,
estimator=weighted_varaint,
orient="v",
col="Type",
join=False,
col_order=type_order_ag)
g4.set_axis_labels("", "Variant Frequency {} CI=95%".format(plus_minus))
g4.set(yscale='log')
g4.set(ylim=(7 * 10**-7, 4 * 10**-3))
g4.set_xticklabels(rotation=90)
# plt.show()
g4.savefig(output_dir + "/Context_label_point_plot", dpi=300)
plt.close()
mutation_g8 = sns.catplot("RNA",
"frac_and_weight",
data=data_filter_ag_replica1,
hue="5`_ADAR_Preference",
palette=mutation_palette(3, adar=True, ag=True),
kind="point",
dodge=True,
estimator=weighted_varaint,
order=rna_order_replica1,
orient="v",
col="Type",
join=False,
col_order=type_order_ag,
hue_order=adar_preference)
mutation_g8.set(yscale="log")
# mutation_g8.fig.suptitle("A>G Mutation trajectories in RV", y=0.99)
mutation_g8.set_axis_labels(
"", "Variant Frequency {} CI=95%".format(plus_minus))
mutation_g8.set(ylim=(1 * 10**-5, 1 * 10**-2))
# plt.show()
mutation_g8.savefig(output_dir + "/ag_ADAR_like_Mutation_col.png", dpi=300)
plt.close()
mutation_g8_box = sns.catplot("RNA",
"Frequency",
data=data_filter_ag_replica1,
hue="5`_ADAR_Preference",
palette=mutation_palette(3,
adar=True,
ag=True),
order=rna_order_replica1,
col="Type",
col_order=type_order_ag,
hue_order=adar_preference,
kind="box")
mutation_g8_box.set(yscale="log")
mutation_g8_box.savefig(output_dir + "/ag_ADAR_like_Mutation_col_box.png",
dpi=300)
data_filter_ag_grouped = data_filter_ag.groupby(
["5`_ADAR_Preference", "label", "Type", "RNA",
"Pos"])["frac_and_weight"].agg(lambda x: weighted_varaint(x))
data_filter_ag_grouped = data_filter_ag_grouped.reset_index()
data_filter_ag_grouped = data_filter_ag_grouped.rename(
columns={"frac_and_weight": "Frequency"})
data_filter_ag_grouped["Frequency"] = data_filter_ag_grouped[
"Frequency"].astype(float)
# data_filter_ag_grouped = data_filter_ag_grouped[data_filter_ag_grouped["RNA"] = "Capsid"]
print(data_filter_ag_grouped.to_string())
data_filter_ag_grouped_silent = data_filter_ag_grouped[
data_filter_ag_grouped["Type"] == "Synonymous"]
# data_filter_ag_grouped_silent = data_filter_ag_grouped_silent[data_filter_ag_grouped_silent["Protein"] != "2A"]
# data_filter_ag_grouped_silent = data_filter_ag_grouped_silent[data_filter_ag_grouped_silent["Protein"] != "3'UTR"]
position_mutation = sns.relplot(
x="Pos",
y="Frequency",
data=data_filter_ag_grouped_silent,
hue="5`_ADAR_Preference",
col="RNA",
col_wrap=2,
style="5`_ADAR_Preference",
palette=mutation_palette(3, adar=True, ag=True),
hue_order=adar_preference,
style_order=["High", "Low", "Intermediate"],
height=4)
position_mutation.set_axis_labels("", "Variant Frequency")
position_mutation.axes.flat[0].set_yscale('symlog', linthreshy=10**-4)
position_mutation.axes.flat[0].set_ylim(10**-4, 10**-1)
plt.savefig(output_dir + "/position_mutation.png", dpi=300)
plt.close()
def main():
# input_dir = "/Volumes/STERNADILABHOME$/volume3/okushnir/AccuNGS/20201008RV-202329127/merged/passages/"
"""Local"""
input_dir = "/Users/odedkushnir/PhD_Projects/After_review/AccuNGS/RV/passages/"
prefix = "inosine_predict_context"
date = datetime.today().strftime("%Y%m%d")
output_dir = input_dir + "{0}_{1}".format(date, prefix)
try:
os.mkdir(output_dir)
except OSError:
print("Creation of the directory %s failed" % output_dir)
else:
print("Successfully created the directory %s " % output_dir)
data_filter = pd.read_pickle(input_dir + prefix + "/data_filter.pkl")
data_filter_ag = pd.read_pickle(input_dir + prefix + "/data_filter_ag.pkl")
data_filter_uc = pd.read_pickle(input_dir + prefix + "/data_filter_uc.pkl")
data_filter["passage"] = data_filter["passage"].astype(int)
data_filter["no_variants"] = np.where(data_filter["Prob"] < 0.95, 0,
data_filter["no_variants"])
data_filter["Read_count"] = data_filter[data_filter["Read_count"] > 10000]
#Plots
label_order = [
"RNA Control\nRND", "RNA Control\nPrimer ID", "p2-1", "p2-2", "p2-3",
"p5-1", "p5-2", "p5-3", "p8-1", "p8-2", "p8-3", "p10-2", "p10-3",
"p12-1", "p12-2", "p12-3"
]
mutation_order = [
"A>G", "U>C", "G>A", "C>U", "A>C", "U>G", "A>U", "U>A", "G>C", "C>G",
"C>A", "G>U"
]
transition_order = ["A>G", "U>C", "G>A", "C>U"]
type_order = ["Synonymous", "Non-Synonymous", "Premature Stop Codon"]
type_order_ag = ["Synonymous", "Non-Synonymous"]
context_order = ["UpA", "ApA", "CpA", "GpA"]
context_order_uc = ["UpU", "UpA", "UpC", "UpG"]
adar_preference = ["High", "Intermediate", "Low"]
plus_minus = u"\u00B1"
# g1 = sns.catplot(x="label", y="frac_and_weight", data=data_filter, hue="Mutation", order=label_order,
# palette="Set2",
# kind="point", dodge=False, hue_order=mutation_order, join=True, estimator=weighted_varaint,
# orient="v")
# g1.set_axis_labels("", "Variant Frequency")
# g1.set_xticklabels(fontsize=9, rotation=45)
# g1.set(yscale='log')
# g1.set(ylim=(10 ** -7, 10 ** -3))
#
# # plt.show()
# g1.savefig(output_dir + "/All_Mutations_point_plot", dpi=300)
# plt.close()
#
# g2 = sns.catplot(x="label", y="frac_and_weight", data=data_filter, hue="Mutation", order=label_order,
# palette=mutation_palette(4), kind="point", dodge=True, hue_order=transition_order, join=False,
# estimator=weighted_varaint,
# orient="v", legend=True)
# g2.set_axis_labels("", "Variant Frequency")
# g2.set(yscale='log', ylim=(10 ** -6, 10 ** -2), xlim=(0, 12, 2))
# # g2.set_yticklabels(fontsize=12)
# g2.set_xticklabels(fontsize=9, rotation=90)
# plt.show()
# g2.savefig("/Users/odedkushnir/Google Drive/Studies/PhD/MyPosters/20190924 GGE/plots/Transition_Mutations_point_plot_RV", dpi=300)
# g2.savefig(output_dir + "/Transition_Mutations_point_plot", dpi=300)
# plt.close()
replica_lst = [1, 2, 3]
for replica in replica_lst:
data_filter_replica = data_filter[data_filter["replica"] == replica]
data_filter_replica["passage"] = data_filter_replica["passage"].astype(
str)
data_filter_replica["passage"] = "p" + data_filter_replica["passage"]
if replica == 2:
data_filter_replica = pd.read_pickle(input_dir + prefix +
"/data_filter.pkl")
data_filter_replica["passage"] = data_filter_replica[
"passage"].astype(int)
data_filter_replica["no_variants"] = np.where(
data_filter_replica["Prob"] < 0.95, 0,
data_filter_replica["no_variants"])
data_filter_replica["Read_count"] = data_filter_replica[
data_filter_replica["Read_count"] > 10000]
data_filter_replica["passage"] = data_filter_replica[
"passage"].astype(str)
data_filter_replica[
"passage"] = "p" + data_filter_replica["passage"]
data_filter_replica["replica"] = np.where(
data_filter_replica["passage"] == "p0", 2,
data_filter_replica["replica"])
data_filter_replica = data_filter_replica[
data_filter_replica["replica"] == replica]
data_filter_replica["passage"] = np.where(
data_filter_replica["passage"] == "p0", "RNA\nControl",
data_filter_replica["passage"])
if replica == 1:
passage_order = ["RNA\nControl", "p2", "p5", "p8", "p12"]
pairs = [(("RNA\nControl", "A>G"), ("RNA\nControl", "G>A")),
(("p2", "A>G"), ("p2", "G>A")),
(("p5", "A>G"), ("p5", "G>A")),
(("p8", "A>G"), ("p8", "G>A")),
(("p12", "A>G"), ("p12", "G>A")),
(("RNA\nControl", "A>G"), ("RNA\nControl", "U>C")),
(("p2", "A>G"), ("p2", "U>C")),
(("p5", "A>G"), ("p5", "U>C")),
(("p8", "A>G"), ("p8", "U>C")),
(("p12", "A>G"), ("p12", "U>C")),
(("RNA\nControl", "A>G"), ("RNA\nControl", "C>U")),
(("p2", "A>G"), ("p2", "C>U")),
(("p5", "A>G"), ("p5", "C>U")),
(("p8", "A>G"), ("p8", "C>U")),
(("p12", "A>G"), ("p12", "C>U"))]
pairs_adar = [(("RNA\nControl", "High\nADAR-like\nA>G"),
("RNA\nControl", "Low\nADAR-like\nA>G")),
(("p2", "High\nADAR-like\nA>G"),
("p2", "Low\nADAR-like\nA>G")),
(("p5", "High\nADAR-like\nA>G"),
("p5", "Low\nADAR-like\nA>G")),
(("p8", "High\nADAR-like\nA>G"),
("p8", "Low\nADAR-like\nA>G")),
(("p12", "High\nADAR-like\nA>G"),
("p12", "Low\nADAR-like\nA>G")),
(("p2", "High\nADAR-like\nU>C"),
("p2", "Low\nADAR-like\nU>C")),
(("p5", "High\nADAR-like\nU>C"),
("p5", "Low\nADAR-like\nU>C")),
(("p8", "High\nADAR-like\nU>C"),
("p8", "Low\nADAR-like\nU>C")),
(("p12", "High\nADAR-like\nU>C"),
("p12", "Low\nADAR-like\nU>C"))]
else:
passage_order = ["RNA\nControl", "p2", "p5", "p8", "p10", "p12"]
pairs = [(("RNA\nControl", "A>G"), ("RNA\nControl", "G>A")),
(("p2", "A>G"), ("p2", "G>A")),
(("p5", "A>G"), ("p5", "G>A")),
(("p8", "A>G"), ("p8", "G>A")),
(("p10", "A>G"), ("p10", "G>A")),
(("p12", "A>G"), ("p12", "G>A")),
(("RNA\nControl", "A>G"), ("RNA\nControl", "U>C")),
(("p2", "A>G"), ("p2", "U>C")),
(("p5", "A>G"), ("p5", "U>C")),
(("p8", "A>G"), ("p8", "U>C")),
(("p10", "A>G"), ("p10", "U>C")),
(("p12", "A>G"), ("p12", "U>C")),
(("RNA\nControl", "A>G"), ("RNA\nControl", "C>U")),
(("p2", "A>G"), ("p2", "C>U")),
(("p5", "A>G"), ("p5", "C>U")),
(("p8", "A>G"), ("p8", "C>U")),
(("p10", "A>G"), ("p10", "C>U")),
(("p12", "A>G"), ("p12", "C>U"))]
pairs_adar = [(("RNA\nControl", "High\nADAR-like\nA>G"),
("RNA\nControl", "Low\nADAR-like\nA>G")),
(("p2", "High\nADAR-like\nA>G"),
("p2", "Low\nADAR-like\nA>G")),
(("p5", "High\nADAR-like\nA>G"),
("p5", "Low\nADAR-like\nA>G")),
(("p8", "High\nADAR-like\nA>G"),
("p8", "Low\nADAR-like\nA>G")),
(("p10", "High\nADAR-like\nA>G"),
("p10", "Low\nADAR-like\nA>G")),
(("p12", "High\nADAR-like\nA>G"),
("p12", "Low\nADAR-like\nA>G")),
(("RNA\nControl", "High\nADAR-like\nU>C"),
("RNA\nControl", "Low\nADAR-like\nU>C")),
(("p2", "High\nADAR-like\nU>C"),
("p2", "Low\nADAR-like\nU>C")),
(("p5", "High\nADAR-like\nU>C"),
("p5", "Low\nADAR-like\nU>C")),
(("p8", "High\nADAR-like\nU>C"),
("p8", "Low\nADAR-like\nU>C")),
(("p10", "High\nADAR-like\nU>C"),
("p10", "Low\nADAR-like\nU>C")),
(("p12", "High\nADAR-like\nU>C"),
("p12", "Low\nADAR-like\nU>C"))]
passage_g = sns.catplot(x="passage",
y="frac_and_weight",
data=data_filter_replica,
hue="Mutation",
order=passage_order,
palette=mutation_palette(4),
kind="point",
dodge=0.5,
hue_order=transition_order,
join=False,
estimator=weighted_varaint,
orient="v",
legend=True)
passage_g.set_axis_labels(
"Passage", "Variant Frequency {} CI=95%".format(plus_minus))
passage_g.set(yscale='log', ylim=(10**-6, 10**-2))
plt.savefig(
output_dir +
"/Transition_Mutations_point_plot_RVB14_replica%s" % str(replica),
dpi=300)
plt.close()
passage_g1 = sns.boxplot(x="passage",
y="Frequency",
data=data_filter_replica,
hue="Mutation",
order=passage_order,
palette=mutation_palette(4),
dodge=True,
hue_order=transition_order)
passage_g1.set_yscale('log')
passage_g1.set_ylim(10**-6, 10**-2)
passage_g1.set(xlabel="Passage", ylabel="Variant Frequency")
annot = Annotator(passage_g1,
pairs,
x="passage",
y="Frequency",
hue="Mutation",
data=data_filter_replica,
order=passage_order,
hue_order=transition_order)
annot.configure(test='t-test_welch',
text_format='star',
loc='outside',
verbose=2,
comparisons_correction="Bonferroni")
annot.apply_test()
file_path = output_dir + "/sts{0}.csv".format(replica)
with open(file_path, "w") as o:
with contextlib.redirect_stdout(o):
passage_g1, test_results = annot.annotate()
plt.legend(bbox_to_anchor=(1.05, 0.5), loc=2, borderaxespad=0.)
plt.tight_layout()
plt.savefig(
output_dir +
"/Transition_Mutations_box_stat_plot_RVB14_replica{0}".format(
replica),
dpi=300)
plt.close()
# data_filter["passage"] = data_filter["passage"].astype(int)
#
#
# g4 = sns.relplot("passage", "frac_and_weight", data=data_filter, hue="Mutation", palette=mutation_palette(4),
# hue_order=transition_order, estimator=weighted_varaint, col="Type", kind="line",
# col_order=type_order)
#
# g4.axes.flat[0].set_yscale('symlog', linthreshy=10 ** -5)
# g4.set_axis_labels("Passage", "Variant Frequency")
# # plt.show()
# g4.savefig(output_dir + "/Time_Transition_Mutations_line_plot", dpi=300)
# plt.close()
"""ADAR preferences"""
data_filter_replica_synonymous = data_filter_replica.loc[
data_filter_replica.Type == "Synonymous"]
# data_filter_synonymous["ADAR_like"] = (data_filter_synonymous.Prev.str.contains('UpA') | data_filter_synonymous.Prev.str.contains('ApA'))
data_filter_replica_synonymous["Mutation"] = np.where(
((data_filter_replica_synonymous["Mutation"] == "A>G") &
(data_filter_replica_synonymous["5`_ADAR_Preference"] == "High")),
"High\nADAR-like\nA>G",
np.where(
((data_filter_replica_synonymous["Mutation"] == "A>G")
& (data_filter_replica_synonymous["5`_ADAR_Preference"]
== "Intermediate")), "Intermediate\nADAR-like\nA>G",
np.where(
((data_filter_replica_synonymous["Mutation"] == "A>G") &
(data_filter_replica_synonymous["5`_ADAR_Preference"]
== "Low")), "Low\nADAR-like\nA>G",
data_filter_replica_synonymous["Mutation"])))
data_filter_replica_synonymous["Mutation_adar"] = np.where(
((data_filter_replica_synonymous["Mutation"] == "U>C") &
(data_filter_replica_synonymous["3`_ADAR_Preference"] == "High")),
"High\nADAR-like\nU>C",
np.where(
((data_filter_replica_synonymous["Mutation"] == "U>C")
& (data_filter_replica_synonymous["3`_ADAR_Preference"]
== "Intermediate")), "Intermediate\nADAR-like\nU>C",
np.where(
((data_filter_replica_synonymous["Mutation"] == "U>C") &
(data_filter_replica_synonymous["3`_ADAR_Preference"]
== "Low")), "Low\nADAR-like\nU>C",
data_filter_replica_synonymous["Mutation"])))
mutation_adar_order = [
"High\nADAR-like\nA>G", "Low\nADAR-like\nA>G",
"High\nADAR-like\nU>C", "Low\nADAR-like\nU>C"
]
# data_filter_replica_synonymous["passage"] = data_filter_replica_synonymous["passage"].astype(str)
# data_filter_replica_synonymous["passage"] = "p" + data_filter_replica_synonymous["passage"]
catplot_adar = sns.catplot(x="passage",
y="frac_and_weight",
data=data_filter_replica_synonymous,
hue="Mutation_adar",
order=passage_order,
palette=mutation_palette(4, adar=True),
kind="point",
dodge=0.5,
hue_order=mutation_adar_order,
join=False,
estimator=weighted_varaint,
orient="v",
legend=True)
catplot_adar.set_axis_labels(
"Passage", "Variant Frequency {} CI=95%".format(plus_minus))
catplot_adar.set(yscale='log')
catplot_adar.set(ylim=(10**-6, 10**-2))
# catplot_adar.set_xticklabels(fontsize=8)
# plt.tight_layout()
plt.savefig(
output_dir +
"/adar_pref_mutation_point_plot_RVB14_replica{0}.png".format(
replica),
dpi=300)
plt.close()
adar_g = sns.boxplot(x="passage",
y="Frequency",
data=data_filter_replica_synonymous,
hue="Mutation_adar",
order=passage_order,
palette=mutation_palette(4, adar=True),
dodge=True,
hue_order=mutation_adar_order)
adar_g.set_yscale('log')
adar_g.set_ylim(10**-6, 10**-1)
adar_g.set(xlabel="Passage", ylabel="Variant Frequency")
annot = Annotator(adar_g,
pairs_adar,
x="passage",
y="Frequency",
hue="Mutation_adar",
data=data_filter_replica_synonymous,
hue_order=mutation_adar_order)
annot.configure(test='t-test_welch',
text_format='star',
loc='outside',
verbose=2,
comparisons_correction="Bonferroni")
annot.apply_test()
file_path = output_dir + "/sts_adar_{0}.csv".format(replica)
with open(file_path, "w") as o:
with contextlib.redirect_stdout(o):
adar_g, test_results = annot.annotate()
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)
plt.tight_layout()
plt.savefig(
output_dir +
"/adar_pref_mutation_box_stat_plot_RVB14_replica{0}".format(
replica),
dpi=300)
plt.close()
def main():
"""RV"""
# input_dir = "/Volumes/STERNADILABHOME$/volume3/okushnir/AccuNGS/20201008RV-202329127/merged/passages"
# Local
input_dir = "/Users/odedkushnir/PhD_Projects/After_review/AccuNGS/RV/passages"
experiment = "passages"
today = datetime.date.today().strftime("%Y%m%d")
output_dir = input_dir + "/{0}_Co_occur_all_{1}".format(today, experiment)
prefix = "/p*"
date = "20201012"
min_coverage = 5000
virus = "RVB14"
q = "q38"
"""1"""
region_lst = [
629, 835, 1621, 2329, 3196, 3634, 3925, 4915, 5170, 5239, 5785, 7165
]
# data_all_passages = creating_co_occur_passages_df(input_dir, experiment, prefix, date, q, region_lst, output_dir)
data_all_passages = creating_co_occur_patients_df(input_dir, experiment,
prefix, date, q,
region_lst, output_dir)
data_all_passages_grouped = data_all_passages.groupby(
["label", "New_Stretch"])["Frequency"].agg(np.mean)
data_all_passages_grouped = data_all_passages_grouped.reset_index()
data_all_passages_grouped = data_all_passages_grouped.rename(
columns={"Frequency": "meanfreq"})
data_all_passages = data_all_passages.merge(data_all_passages_grouped,
how="left",
on=["label", "New_Stretch"])
# label_order = ["p2-1", "p2-2", "p2-3", "p5-1", "p5-2", "p5-3", "p8-1", "p8-2", "p8-3", "p10-2", "p10-3", "p12-1",
# "p12-2", "p12-3"]
label_order = ["p2-3", "p5-3", "p8-3", "p10-3", "p12-3"]
style_order = [
"No editing context", "ADAR (antisense)", "ADAR (sense)", "APOBEC3F"
]
style_adar = ["ADAR (antisense)", "ADAR (sense)"]
markers = {
"No editing context": "o",
"ADAR (antisense)": "<",
"ADAR (sense)": ">",
"APOBEC3F": "*"
}
markers_adar = {"ADAR (antisense)": "<", "ADAR (sense)": ">"}
hue_order = ["A>G", "U>C", "G>A", "C>U"]
hue_order_adar = ["A>G", "U>C"]
file_name = "Passages_Editing_context"
plot_editing_context_plot(data_all_passages,
output_dir,
file_name,
label_order,
style_order=style_order,
markers=markers,
hue_order=hue_order,
mutation_palette=mutation_palette(4),
experiment=experiment,
col_wrap=3)
data_all_passages = data_all_passages[data_all_passages["ADAR"] != "No"]
file_name = "Passages_ADAR_context"
plot_editing_context_plot(data_all_passages,
output_dir,
file_name,
label_order,
style_order=style_adar,
markers=markers_adar,
hue_order=hue_order_adar,
mutation_palette=mutation_palette(2,
adar=True,
ag=True,
uc=True),
experiment=experiment,
col_wrap=3)
# """2"""
# # data_all_passages = pd.read_pickle(output_dir + "/all_co_occur_protein.pkl")
# df_co_occur_new = grouped_co_occur(data_all_passages, input_dir, experiment, output_dir, q)
#
# g1 = sns.boxenplot(x="Stretch_Type_Non-Synonymous", y="Stretch_Freq", data=df_co_occur_new,
# order=[False, True], color="0.8")
# g1 = sns.stripplot(x="Stretch_Type_Non-Synonymous", y="Stretch_Freq", data=df_co_occur_new,
# order=[False, True], palette=[sns.color_palette("muted")[3], sns.color_palette("muted")[4]])
# old_statannot.add_stat_annotation(g1, data=df_co_occur_new, x="Stretch_Type_Non-Synonymous", y="Stretch_Freq",
# boxPairList=[(False, True)], test='Mann-Whitney', textFormat='star',
# loc='outside', verbose=2,
# order=[False, True])
# g1.set(xticklabels=("Synonymous\nStretches", "Stretches\nWith Non-Synonymous"))
# g1.set(xlabel="")
# g1.set(ylabel = "Stretch Frequency")
# g1.set(ylim= (0, 0.007))
# plt.tight_layout()
# plt.savefig(output_dir + "/Frequencies_of_Stertch_Type_Non-Synonymous", dpi=300)
# plt.close()
# df_adar_path = "/Volumes/STERNADILABHOME$/volume3/okushnir/AccuNGS/20201008RV-202329127/merged/passages/" \
# "inosine_predict_context/data_filter.csv"
# """3"""
# # df_co_occur_new = pd.read_pickle(output_dir + "/all_co_occur_grouped.pkl")
# df_regression = regression_stretches(df_co_occur_new, output_dir, df_adar_path, experiment, min_coverage)
# plot_regression(df_regression, output_dir)
# """4"""
# # df_regression = pd.read_pickle(output_dir + "/all_co_occur_grouped_adar_preferences.pkl")
# plot_regression_stretches_adar_preferences(df_regression, output_dir)
"""RV-Capsid_Free"""
# input_dir = "/Volumes/STERNADILABHOME$/volume3/okushnir/AccuNGS/20201008RV-202329127/merged/capsid"
# experiment = "capsid"
# output_dir = input_dir + "/20201129Co_occur_all_%s" % experiment
# prefix = "/*_3*"
# min_coverage = 5000
# virus = "RVB14"
# date = "20201012"
# q = "q38"
# """1"""
# region_lst = [629, 835, 1621, 2329, 3196, 3634, 3925, 4915, 5170, 5239, 5785, 7165]
# # data_all_capsid = creating_co_occur_passages_df(input_dir, experiment, prefix, date, q, region_lst, output_dir)
# data_all_capsid = creating_co_occur_patients_df(input_dir, experiment, prefix, date, q, region_lst, output_dir)
# data_all_capsid["label"] = np.where(data_all_capsid["label"] == "Free-31-Amicon", "Free #1",
# data_all_capsid["label"])
# data_all_capsid["label"] = np.where(data_all_capsid["label"] == "Free-32-Ultra", "Free #2",
# data_all_capsid["label"])
# data_all_capsid["label"] = np.where(data_all_capsid["label"] == "Free-33-Ultra", "Free #3",
# data_all_capsid["label"])
# data_all_capsid["label"] = np.where(data_all_capsid["label"] == "Capsid-31-Amicon", "Capsid #1",
# data_all_capsid["label"])
# data_all_capsid["label"] = np.where(data_all_capsid["label"] == "Capsid-33-Ultra", "Capsid #3",
# data_all_capsid["label"])
# label_order = ["Free #1", "Free #2", "Free #3", "Capsid #1", "Capsid #3"]
# style_order = ["No editing context", "ADAR (antisense)", "ADAR (sense)", "APOBEC3F"]
# style_adar = ["ADAR (antisense)", "ADAR (sense)"]
# markers = {"No editing context": "o", "ADAR (antisense)": "<", "ADAR (sense)": ">", "APOBEC3F": "*"}
# markers_adar = {"ADAR (antisense)": "<", "ADAR (sense)": ">"}
# hue_order = ["A>G", "U>C", "G>A", "C>U"]
# hue_order_adar = ["A>G", "U>C"]
# file_name = "Capsid_Editing_context"
# plot_editing_context_plot(data_all_capsid, output_dir, file_name, label_order, col_wrap=3, style_order=style_order,
# markers=markers, hue_order=hue_order, mutation_palette=mutation_palette(4))
# data_all_capsid = data_all_capsid[data_all_capsid["ADAR"] != "No"]
# file_name = "Capsid_ADAR_context"
# plot_editing_context_plot(data_all_capsid, output_dir, file_name, label_order, col_wrap=3, style_order=style_adar,
# markers=markers_adar, hue_order=hue_order_adar,
# mutation_palette=mutation_palette(2, adar=True, ag=True, uc=True))
# """2"""
# # data_all_capsid = pd.read_pickle(output_dir + "/all_co_occur_protein.pkl")
# # df_co_occur_new = grouped_co_occur(data_all_capsid, input_dir, experiment, output_dir, q)
# #
# # Plots
# df_co_occur_new = pd.read_pickle(output_dir + "/all_co_occur_grouped.pkl")
# g1 = sns.catplot(x="RNA", y="Stretch_Freq", data=df_co_occur_new, order=["Capsid", "Free"], kind="strip",
# col="replica", palette=[sns.color_palette("muted")[3], sns.color_palette("muted")[4]],
# legend=True)
# g1.set(yscale='log')
# g1.set(ylim=(10**-5, 10**-1))
# plt.tight_layout()
# plt.savefig(output_dir + "/Frequencies_of_Stertch", dpi=300)
# plt.close()
# stretch_g1 = sns.catplot(x="Stretch_Type_Non-Synonymous", y="Stretch_Freq", data=df_co_occur_new,
# order=[False, True], hue="RNA", hue_order=["Capsid", "Free"], kind="strip", col="replica",
# palette=[sns.color_palette()[2], sns.color_palette()[6]],
# sharey=True)
# stretch_g1.set(xticklabels=("Synonymous\nStretches", "Stretches\nWith Non-Synonymous"))
# stretch_g1.set(xlabel="")
# stretch_g1.set(ylabel = "Stretch Frequency")
# stretch_g1.set(ylim= (0, 0.007))
#
# plt.tight_layout()
#
# # g1 = sns.boxenplot(x="Stretch_Type_Non-Synonymous", y="Stretch_Freq", data=df_co_occur_new,
# # order=[False, True], color="0.8", hue="RNA", hue_order=["Capsid", "Free"])
# # g1 = sns.stripplot(x="Stretch_Type_Non-Synonymous", y="Stretch_Freq", data=df_co_occur_new,
# # order=[False, True], hue="RNA", hue_order=["Capsid", "Free"], palette=[sns.color_palette("muted")[3], sns.color_palette("muted")[4]])
# # # old_statannot.add_stat_annotation(g1, data=df_co_occur_new, x="Stretch_Type_Non-Synonymous", y="Stretch_Freq",
# # # boxPairList=[(False, True)], test='Mann-Whitney', textFormat='star',
# # # loc='outside', verbose=2,
# # # order=[False, True])
# # g1.set(xticklabels=("Synonymous\nStretches", "Stretches\nWith Non-Synonymous"))
# # g1.set(xlabel="")
# # g1.set(ylabel = "Stretch Frequency")
# plt.savefig(output_dir + "/Frequencies_of_Stertch_Type_Non-Synonymous_replicas", dpi=300)
#
# """3"""
# df_adar_path = "/Volumes/STERNADILABHOME$/volume3/okushnir/AccuNGS/20201008RV-202329127/merged/capsid/" \
# "inosine_predict_context/data_filter.csv"
#
# # df_co_occur_new = pd.read_pickle(output_dir + "/all_co_occur_grouped.pkl")
# df_regression = regression_stretches(df_co_occur_new, output_dir, df_adar_path, experiment, min_coverage)
# """4"""
# # df_regression = pd.read_pickle(output_dir + "/all_co_occur_grouped_adar_preferences.pkl")
# plot_capsid_free_plot(df_regression, output_dir)
"""RV-Patients"""
# input_dir = "/Volumes/STERNADILABHOME$/volume3/okushnir/AccuNGS/20201008RV-202329127/merged/patients"
# experiment = "patients"
# output_dir = input_dir + "/20201202Co_occur_%s" % experiment
# prefix = "/Patient_*"
# min_coverage = 5000
# virus = "RVA"
# date = "20201124"
# q = "q30_consensusX7"
#
#
# region_lst = [503, 709, 1495, 2197, 3094, 3523, 3808, 4771, 5005, 5068, 5617, 6728]
# data_all_patients = creating_co_occur_patients_df(input_dir, experiment, prefix, date, q, region_lst, output_dir)
# label_order = ["Patient-1", "Patient-4", "Patient-5", "Patient-9", "Patient-16", "Patient-17", "Patient-20"]
# style_order = ["No editing context", "ADAR (antisense)", "ADAR (sense)", "APOBEC3F"]
# style_adar = ["ADAR (antisense)", "ADAR (sense)"]
# markers = {"No editing context": "o", "ADAR (antisense)": "<", "ADAR (sense)": ">", "APOBEC3F": "*"}
# markers_adar = {"ADAR (antisense)": "<", "ADAR (sense)": ">"}
# hue_order = ["A>G", "U>C", "G>A", "C>U"]
# hue_order_adar = ["A>G", "U>C"]
# file_name = "Patient_Editing_context"
# plot_editing_context_plot(data_all_patients, output_dir, file_name, label_order, col_wrap=3, style_order=style_order
# , markers=markers, hue_order=hue_order, mutation_palette=mutation_palette(4))
# data_all_patients = data_all_patients[data_all_patients["ADAR"] != "No"]
# file_name = "Patient_ADAR_context"
# plot_editing_context_plot(data_all_patients, output_dir, file_name, label_order, col_wrap=3, style_order=style_adar,
# markers=markers_adar, hue_order=hue_order_adar,
# mutation_palette=mutation_palette(2, adar=True, ag=True, uc=True))
#
# # data_all_patients = pd.read_pickle(output_dir + "/all_co_occur_protein.pkl")
#
# df_co_occur_new = grouped_co_occur(data_all_patients, input_dir,experiment, output_dir, q=q.split("_")[0])
# plot_patient_plot(df_co_occur_new, output_dir, "Frequencies_of_label_Stertch_Type_Non-Synonymous")
# df_adar_path = "/Volumes/STERNADILABHOME$/volume3/okushnir/AccuNGS/20201008RV-202329127/merged/patients/" \
# "inosine_predict_context/data_filter.csv"
# df_co_occur_new = pd.read_pickle(output_dir + "/all_co_occur_grouped.pkl")
# df_regression = regression_stretches(df_co_occur_new, output_dir, df_adar_path, experiment, min_coverage)
# df_regression = pd.read_pickle(output_dir + "/all_co_occur_grouped_adar_preferences.pkl")
# plot_patient_plot(df_regression, output_dir)
"""CV"""