def logo_rbs_from_gms2_mod_file(pd_figures, pf_mod, title=""):
# type: (str, str, str) -> None
mod = GMS2Mod.init_from_file(pf_mod)
mm = MotifModel(mod.items["RBS_MAT"], mod.items["RBS_POS_DISTR"])
non = GMS2Noncoding(mod.items["NON_MAT"])
import matplotlib.pyplot as plt
fig, axes = plt.subplots(1, 2)
import logomaker as lm
lm.Logo(lm.transform_matrix(mm.pwm_to_df(),
from_type="probability",
to_type="information",
background=non.pwm_to_array(0)),
ax=axes[0])
axes[0].set_title(title)
axes[0].set_ylim(0, 2)
df_spacer = pd.DataFrame({
"Distance from start": range(len(mm._spacer)),
"Probability": mm._spacer
})
sns.lineplot(df_spacer,
"Distance from start",
"Probability",
ax=axes[1],
figure_options=FigureOptions(ylim=[0, 0.4]))
plt.tight_layout()
plt.savefig(next_name(pd_figures))
plt.show()
def plot_sensitivities_vs_num_candidates(sensitiviies_func, max_candidates,
sen_a, sen_b):
# type: (Dict[str, Callable], int, float, float) -> None
list_entries = list()
for i in range(1, max_candidates + 1):
curr_sensitivities = {
name: sensitiviies_func[name](i, sen_a, sen_b)
for name in sensitiviies_func.keys()
}
list_entries.append({"Number of candidates": i, **curr_sensitivities})
df = pd.DataFrame(list_entries)
conditions = sorted(list(sensitiviies_func.keys()))
df_stacked = stack_columns_as_rows(df, conditions, "Probability",
conditions, "Condition")
sns.lineplot(df_stacked,
"Number of candidates",
"Probability",
hue="Condition")
def analyze_by_support(df, pd_work, fn_prefix, tag):
# type: (pd.DataFrame, str, str, str) -> None
list_df = list()
for index in df.index:
curr_df = pd.DataFrame(df.at[index, "by_support_{}".format(tag)])
curr_df["Genome"] = df.at[index, "Genome"]
if df.at[index, "Genome"] in {"A. pernix", "Synechocystis"}:
continue
list_df.append(curr_df)
df_acc = pd.concat(list_df)
sns.lineplot(
df_acc,
"Min Support",
"Percentage 3p match: Verified from {}".format(tag),
hue="Genome",
figure_options=FigureOptions(
title="Percentage of verified genes predicted\nby {}".format(tag),
ylabel="Percentage",
save_fig=next_name(pd_work),
ylim=[None, 100.5]))
sns.lineplot(
df_acc,
"Min Support",
"Percentage 5p-3p match: Verified from {}".format(tag),
hue="Genome",
figure_options=FigureOptions(
title="Percentage of predicted {} genes\nwith correct 5' end".
format(tag),
ylabel="Percentage of 5p-3p match",
save_fig=next_name(pd_work),
ylim=[90, 100.5]))
def main(env, args):
# type: (Environment, argparse.Namespace) -> None
gil = GenomeInfoList.init_from_file(args.pf_genome_list)
prl_options = ParallelizationOptions.init_from_dict(env, vars(args))
if not prl_options["use-pbs"]:
df = relative_entropy_analysis(env, gil, prl_options)
else:
pbs = PBS(env,
prl_options,
splitter=split_genome_info_list,
merger=merge_identity)
list_df = pbs.run(data={"gil": gil},
func=relative_entropy_analysis,
func_kwargs={
"env": env,
"prl_options": prl_options
})
df = pd.concat(list_df, ignore_index=True, sort=False)
df.to_csv(os_join(env["pd-work"], "summary.csv"), index=False)
pd_figures = os_join(env["pd-work"], "summary_figures")
mkdir_p(pd_figures)
sns.scatterplot(df,
"Percent",
"Error",
figure_options=FigureOptions(
ylim=[0, 20], save_fig=next_name(pd_figures)))
sns.lineplot(df,
"RE",
"Error",
hue="Genome",
figure_options=FigureOptions(ylim=[0, 20],
save_fig=next_name(pd_figures)))
sns.lineplot(df,
"RE Motif",
"Error",
hue="Genome",
figure_options=FigureOptions(ylim=[0, 20],
save_fig=next_name(pd_figures)))
sns.lineplot(df,
"RE Spacer",
"Error",
hue="Genome",
figure_options=FigureOptions(ylim=[0, 20],
save_fig=next_name(pd_figures)))
sns.scatterplot(
df,
"RE Motif",
"RE Spacer",
hue="Genome",
identity=True,
figure_options=FigureOptions(save_fig=next_name(pd_figures)))
sns.lmplot(df,
"Percent",
"Error",
hue="Genome",
figure_options=FigureOptions(ylim=[0, 20],
save_fig=next_name(pd_figures)))
sns.lmplot(df,
"RE",
"Error",
hue="Genome",
figure_options=FigureOptions(ylim=[0, 20],
save_fig=next_name(pd_figures)))
sns.lmplot(df,
"RE Motif",
"Error",
hue="Genome",
figure_options=FigureOptions(ylim=[0, 20],
save_fig=next_name(pd_figures)))
sns.lmplot(df,
"RE Spacer",
"Error",
hue="Genome",
figure_options=FigureOptions(ylim=[0, 20],
save_fig=next_name(pd_figures)))
sns.lmplot(df,
"Percent",
"RE",
hue="Genome",
figure_options=FigureOptions(save_fig=next_name(pd_figures)))
def one_dim_Kimura_accuracy(env, df_all, num_steps=20):
# type: (Environment, pd.DataFrame, int) -> None
import matplotlib.pyplot as plt
pd_work = env["pd-work"]
ancestors = sorted(list(set(df_all["Ancestor"])))
# fig, axes = plt.subplots(2, math.ceil(len(ancestors) / 2), sharex=True, sharey=True)
# min_x = min(df_all["Average-Kimura"])
# max_x = max(df_all["Average-Kimura"]) + 0.000000001
# ss_x = (max_x - min_x) / float(num_steps)
#
# list_df = list()
# axis_idx = 0
# for ancestor, df in df_all.groupby("Ancestor", as_index=False):
# # ax = axes.ravel()[axis_idx]
# # axis_idx += 1
#
#
#
#
#
# import numpy as np
# gms2_eq_sbsp_and_ncbi = np.zeros(num_steps, dtype=float)
# gms2_eq_sbsp_eq_ncbi = np.zeros(num_steps, dtype=float)
#
# df_gms2_eq_sbsp_and_ncbi = (df["GMS2=SBSP"]) & (df["NCBI"])
# df_gms2_eq_sbsp_eq_ncbi = (df["GMS2=SBSP=NCBI"])
#
# for index in df.index:
#
# x_val = df.at[index, "Average-Kimura"]
#
# x_pos = int((x_val-min_x) / ss_x)
#
# gms2_eq_sbsp_and_ncbi[x_pos] += 1 if df.at[index, "GMS2=SBSP"] and df.at[index, "NCBI"] else 0
# gms2_eq_sbsp_eq_ncbi[x_pos] += 1 if df.at[index, "GMS2=SBSP=NCBI"] else 0
#
# accuracy = np.divide(gms2_eq_sbsp_eq_ncbi, gms2_eq_sbsp_and_ncbi)
# # accuracy = np.flip(accuracy, 0)
#
#
# xticks = list(range(0, num_steps))
#
# l_x = np.arange(min_x, max_x, ss_x)
# xticklabels = [round(l_x[i], 2) for i in xticks]
# # g = seaborn.heatmap(accuracy.transpose(), vmin=0, vmax=1, xticklabels=xticklabels, yticklabels=yticklabels, ax=ax,
# # cbar=True)
#
# # g = seaborn.lineplot(xticklabels, accuracy, ax=ax, label=ancestor)
#
# # cbar=g.cbar
#
# # g.set_xticks(xticks)
#
# curr_df = pd.DataFrame({
# "Average-Kimura": xticklabels,
# "Accuracy": accuracy,
# "Number-of-queries": gms2_eq_sbsp_and_ncbi
# })
# curr_df["Ancestor"] = ancestor
# list_df.append(curr_df)
#
# # g.set_xlabel("Min Kimura")
# # g.set_ylabel("Max Kimura")
# # g.set_title(ancestor)
#
# df = pd.concat(list_df) # type: pd.DataFrame
df = bin_data_one_d(env, df_all, "Average-Kimura", num_steps)
sns.lineplot(df,
"Average-Kimura",
"Accuracy",
hue="Ancestor",
figure_options=FigureOptions(save_fig=next_name(pd_work), ),
sns_kwargs={"palette": CM.get_map("ancestor")})
sns.lineplot(df,
"Average-Kimura",
"Number-of-queries",
hue="Ancestor",
figure_options=FigureOptions(save_fig=next_name(pd_work), ),
sns_kwargs={"palette": CM.get_map("ancestor")})
total_per_ancestor = {
ancestor: (df["Ancestor"].isin({ancestor})).sum()
for ancestor in ancestors
}
df["Percentage-of-queries"] = 0
df["Cumulative-percentage-of-queries"] = 0
df.reset_index(inplace=True)
for ancestor, df_group in df.groupby(
"Ancestor", as_index=False): # type: str, pd.DataFrame
df_group.sort_values("Average-Kimura", inplace=True)
index = df_group.index
prev = 0
total = df_group["Number-of-queries"].sum()
df.loc[index, "Percentage-of-queries"] = 100 * df.loc[
index, "Number-of-queries"] / float(total)
for i in index:
df.loc[i, "Cumulative-percentage-of-queries"] = prev + df.loc[
i, "Percentage-of-queries"]
prev = df.loc[i, "Cumulative-percentage-of-queries"]
fig, ax = plt.subplots(figsize=(8, 4))
sns.lineplot(df,
"Average-Kimura",
"Percentage-of-queries",
hue="Ancestor",
figure_options=FigureOptions(save_fig=next_name(pd_work),
ylabel="Percentage of queries",
xlabel="Average Kimura"),
ax=ax,
show=True,
legend_loc="best",
sns_kwargs={"palette": CM.get_map("ancestor")})
sns.lineplot(df,
"Average-Kimura",
"Cumulative-percentage-of-queries",
hue="Ancestor",
figure_options=FigureOptions(save_fig=next_name(pd_work), ),
sns_kwargs={"palette": CM.get_map("ancestor")})
# standard dev
df = bin_data_one_d(env, df_all[df_all["Support"] > 2], "Std-Kimura",
num_steps)
sns.lineplot(df,
"Std-Kimura",
"Accuracy",
hue="Ancestor",
figure_options=FigureOptions(save_fig=next_name(pd_work), ),
sns_kwargs={"palette": CM.get_map("ancestor")})
sns.lineplot(df,
"Std-Kimura",
"Number-of-queries",
hue="Ancestor",
figure_options=FigureOptions(save_fig=next_name(pd_work), ),
sns_kwargs={"palette": CM.get_map("ancestor")})
total_per_ancestor = {
ancestor: (df["Ancestor"].isin({ancestor})).sum()
for ancestor in ancestors
}
df["Percentage-of-queries"] = 0
df["Cumulative-percentage-of-queries"] = 0
df.reset_index(inplace=True)
for ancestor, df_group in df.groupby(
"Ancestor", as_index=False): # type: str, pd.DataFrame
df_group.sort_values("Std-Kimura", inplace=True)
index = df_group.index
prev = 0
total = df_group["Number-of-queries"].sum()
df.loc[index, "Percentage-of-queries"] = 100 * df.loc[
index, "Number-of-queries"] / float(total)
for i in index:
df.loc[i, "Cumulative-percentage-of-queries"] = prev + df.loc[
i, "Percentage-of-queries"]
prev = df.loc[i, "Cumulative-percentage-of-queries"]
sns.lineplot(df,
"Std-Kimura",
"Percentage-of-queries",
hue="Ancestor",
figure_options=FigureOptions(save_fig=next_name(pd_work), ),
sns_kwargs={"palette": CM.get_map("ancestor")})
sns.lineplot(df,
"Std-Kimura",
"Cumulative-percentage-of-queries",
hue="Ancestor",
figure_options=FigureOptions(save_fig=next_name(pd_work), ),
sns_kwargs={"palette": CM.get_map("ancestor")})
def viz_summary_per_gcfid_per_step(env, df):
# type: (Environment, pd.DataFrame) -> None
pd_work = env['pd-work']
list_df = list()
for gcfid, df_group in df.groupby("GCFID", as_index=False):
df.loc[df_group.index, "Total SBSP"] = df.loc[df_group.index,
"SBSP"].sum()
df.loc[df_group.index, "Total GMS2"] = df.loc[df_group.index,
"GMS2"].sum()
df.loc[df_group.index, "Total GMS2=SBSP"] = df.loc[df_group.index,
"GMS2=SBSP"].sum()
tag = None
for step in ["A", "B", "C"]:
if tag is None:
tag = step
else:
tag += "+" + step
df_summary_per_gcfid = get_summary_per_gcfid(
df[df["Predicted-at-step"] <= step])
df_summary_per_gcfid["SBSP Step"] = tag
list_df.append(df_summary_per_gcfid)
df_per_gcfid_per_step = pd.concat(list_df, sort=False)
import matplotlib.pyplot as plt
# fig, ax = plt.subplots()
#
# sns.lineplot(df_per_gcfid_per_step, "SBSP Step", "SBSP", hue="GCFID", ax=ax,
# sns_kwargs={"palette": CM.get_map("verified")},
# legend=False
# )
# for l in ax.lines:
# l.set_linestyle("--")
#
# ax2 = ax.twinx()
# sns.lineplot(df_per_gcfid_per_step, "SBSP Step", "Sen(SBSP,NCBI)", hue="GCFID", ax=ax2,
# sns_kwargs={"palette": CM.get_map("verified")},)
#
# fo = FigureOptions(
# xlabel="SBSP Step",
# ylabel="Percentage",
# # ylim=[0, 105],
# save_fig=next_name(env["pd-work"])
# )
# FigureOptions.set_properties_for_axis(ax, fo)
# plt.subplots_adjust(bottom=0.2)
# handles, labels = ax.get_legend_handles_labels()
# ax.legend(handles=handles[1:], labels=labels[1:],
# loc="lower center", ncol=4, bbox_to_anchor=(0.5, -0.25))
#
# plt.savefig(fo.save_fig)
# plt.show()
fig, axes = plt.subplots(3, 2, sharex="all", sharey="row")
ax = axes[:, 0]
sns.lineplot(df_per_gcfid_per_step,
"SBSP Step",
"Sen(SBSP,NCBI)",
hue="GCFID",
ax=ax[0],
sns_kwargs={"palette": CM.get_map("verified")},
legend=False,
figure_options=FigureOptions(
ylabel="Sensitivity",
ylim=[85, 105],
))
sns.lineplot(df_per_gcfid_per_step,
"SBSP Step",
"Cov(SBSP,NCBI)",
hue="GCFID",
ax=ax[1],
sns_kwargs={"palette": CM.get_map("verified")},
legend=False,
figure_options=FigureOptions(ylabel="Percent of Genes",
ylim=[0, None]))
sns.lineplot(df_per_gcfid_per_step,
"SBSP Step",
"SBSP",
hue="GCFID",
ax=ax[2],
sns_kwargs={"palette": CM.get_map("verified")},
legend=False,
figure_options=FigureOptions(ylabel="Number of Genes",
ylim=[0, None]))
fig.align_ylabels(ax)
# plt.savefig(next_name(env["pd-work"]))
# plt.show()
# fig, ax = plt.subplots(3, 1, sharex="all")
ax = axes[:, 1]
sns.lineplot(df_per_gcfid_per_step,
"SBSP Step",
"Sen(GMS2=SBSP,NCBI)",
hue="GCFID",
ax=ax[0],
sns_kwargs={"palette": CM.get_map("verified")},
legend=False,
figure_options=FigureOptions(
ylabel="Sensitivity",
ylim=[85, 105],
))
sns.lineplot(df_per_gcfid_per_step,
"SBSP Step",
"Cov(GMS2=SBSP,NCBI)",
hue="GCFID",
ax=ax[1],
sns_kwargs={"palette": CM.get_map("verified")},
legend=False,
figure_options=FigureOptions(ylabel="Percent of Genes",
ylim=[0, None]))
sns.lineplot(df_per_gcfid_per_step,
"SBSP Step",
"GMS2=SBSP",
hue="GCFID",
ax=ax[2],
sns_kwargs={"palette": CM.get_map("verified")},
figure_options=FigureOptions(ylabel="Number of Genes",
ylim=[0, None]))
ax[2].get_legend().remove()
fig.align_ylabels(ax)
for ax in axes.ravel():
ax.set_xlabel("Steps")
axes[0][0].set_title("SBSP")
axes[0][1].set_title("GMS2=SBSP")
fig.subplots_adjust(bottom=0.21)
# handles, labels = ax.get_legend_handles_labels()
# fig.legend(handles=handles[1:], labels=labels[1:], loc="lower center", ncol=4)#, bbox_to_anchor=(0.5, -0.25))
handles, labels = ax.get_legend_handles_labels()
labels[0] = "Genome"
fig.legend(handles=handles, labels=labels, loc="lower center",
ncol=3) #, bbox_to_anchor=(0.5, -0.25))
plt.savefig(next_name(env["pd-work"]))
plt.show()
# three plots
for gcfid, df_group in df.groupby("GCFID", as_index=False):
df.loc[df_group.index,
"Total SBSP"] = ((df_group["SBSP"]) & (df_group["NCBI"])).sum()
df.loc[df_group.index,
"Total GMS2"] = ((df_group["GMS2"]) & (df_group["NCBI"])).sum()
df.loc[df_group.index, "Total GMS2=SBSP"] = ((df_group["GMS2=SBSP"]) &
(df_group["NCBI"])).sum()
df_all = get_summary_per_gcfid(df)
print(df_all[[
"GCFID", "NCBI", "Sen(SBSP,NCBI)", "Sen(GMS2,NCBI)",
"Sen(GMS2=SBSP,NCBI)"
]].to_string(index=False))
print(df_all[[
"GCFID", "NCBI", "Cov2(SBSP,NCBI)", "Cov2(GMS2,NCBI)",
"Cov2(GMS2=SBSP,NCBI)"
]].to_string(index=False))
import sys
sys.exit()
def viz_per_genome(env, df):
# type: (Environment, pd.DataFrame) -> None
df_grp = df.groupby(["Genome", "Ancestor"], as_index=False).mean()
sns.catplot(df_grp,
"Ancestor",
"BLAST",
figure_options=FigureOptions(save_fig=next_name(
env["pd-work"]),
xlabel="Clade",
ylabel="Number of BLASTp Hits"),
sns_kwargs={"palette": CM.get_map("ancestor")})
# list_grp = list()
# for _, df_grp in df.groupby("Genome", as_index=False):
# indices = df_grp.index
#
# list_grp.append({
# "Genome": df.at[indices[0], "Genome"],
# "Ancestor": df.at[indices[0], "Ancestor"],
# "= 0": len(df_grp[df_grp["BLAST"] == 0]),
# **{
# f"< {x}": len(df_grp[df_grp["BLAST"] < x]) for x in [5, 10, 20, 50, 100, 500, 1000, 5000, 10000]
# },
# "> 10000": len(df_grp[df_grp["BLAST"] > 10000])
# })
#
# df_grp = pd.DataFrame(list_grp)
# sns.catplot(df_grp, "Ancestor", "= 0")
# sns.catplot(df_grp, "Ancestor", "< 5")
# sns.catplot(df_grp, "Ancestor", "< 50")
# sns.catplot(df_grp, "Ancestor", "< 100")
# plots
# 1) x: number of queries with < x targets
# compute per genome, the % of queries with hits <= 0, 5, 10, 20, 40, 80, 160, ... 240 580 1160, ...
# plot
list_entries = list()
for _, df_grp in df.groupby("Genome", as_index=False):
indices = df_grp.index
genome = df.at[indices[0], "Genome"]
ancestor = df.at[indices[0], "Ancestor"]
total_queries = len(df_grp)
curr = 0
for n in range(40):
list_entries.append({
"Genome":
genome,
"Ancestor":
ancestor,
"x":
curr,
"y":
100 * len(df_grp[df_grp["BLAST"] < curr]) / total_queries
})
# if list_entries[-1]["y"] == 100:
# break
if curr == 0:
curr = 5
else:
curr *= 1.2
df_tmp = pd.DataFrame(list_entries)
SMALL_SIZE = 16
MEDIUM_SIZE = 22
BIGGER_SIZE = 24
matplotlib.rcParams.update({
# "pgf.texsystem": "pdflatex",
'font.family': 'serif',
'text.usetex': True,
'pgf.rcfonts': False,
'font.size': SMALL_SIZE, # controls default text sizes
'axes.titlesize': SMALL_SIZE, # fontsize of the axes title
'axes.labelsize': MEDIUM_SIZE, # fontsize of the x and y labels
'xtick.labelsize': SMALL_SIZE, # fontsize of the tick labels
'ytick.labelsize': SMALL_SIZE, # fontsize of the tick labels
'legend.fontsize': 12, # legend fontsize
'figure.titlesize': BIGGER_SIZE, # fontsize of the figure title
})
sns.lineplot(df_tmp,
"x",
"y",
hue="Ancestor",
figure_options=FigureOptions(
xlabel="Number of BLASTp hits",
ylabel="Cumulative percentage of queries (per genome)",
save_fig=next_name(env["pd-work"]),
),
legend_loc="best",
legend_title="",
legend_ncol=2,
sns_kwargs={
"ci": "sd",
"palette": CM.get_map("ancestor")
})
sns.lineplot(df_tmp,
"y",
"x",
hue="Ancestor",
figure_options=FigureOptions(
ylabel="Number of BLASTp hits",
xlabel="Cumulative percentage of queries (per genome)",
save_fig=next_name(env["pd-work"]),
),
legend_loc="best",
legend_title="",
legend_ncol=2,
sns_kwargs={
"ci": "sd",
"palette": CM.get_map("ancestor")
})
SMALL_SIZE = 14
MEDIUM_SIZE = 18
BIGGER_SIZE = 20
matplotlib.rcParams.update({
# "pgf.texsystem": "pdflatex",
'font.family': 'serif',
'text.usetex': True,
'pgf.rcfonts': False,
'font.size': SMALL_SIZE, # controls default text sizes
'axes.titlesize': SMALL_SIZE, # fontsize of the axes title
'axes.labelsize': MEDIUM_SIZE, # fontsize of the x and y labels
'xtick.labelsize': SMALL_SIZE, # fontsize of the tick labels
'ytick.labelsize': SMALL_SIZE, # fontsize of the tick labels
'legend.fontsize': 12, # legend fontsize
'figure.titlesize': BIGGER_SIZE, # fontsize of the figure title
})
fig, axes = plt.subplots(2, 2, sharex="all", sharey="all")
ancestors = sorted(set(df["Ancestor"]))
for anc, ax in zip(ancestors, axes.ravel()):
df_anc = df_tmp[df_tmp["Ancestor"] == anc]
sns.lineplot(df_anc[df_anc["x"] <= 40],
"x",
"y",
hue="Ancestor",
legend=None,
ax=ax,
sns_kwargs={
"ci": "sd",
"palette": CM.get_map("ancestor")
})
ax.set_title(anc)
ax.set_xlabel("")
ax.set_ylabel("")
figure_options = FigureOptions(
xlabel="Number of BLASTp hits",
ylabel="Cumulative percentage of\nqueries (per genome)",
save_fig=next_name(env["pd-work"]),
)
fig.add_subplot(111, frameon=False)
# # hide tick and tick label of the big axes
plt.tick_params(top=False,
bottom=False,
left=False,
right=False,
which="both",
labelbottom=False,
labeltop=False,
labelleft=False,
labelright=False)
plt.xlabel(figure_options.xlabel, labelpad=30)
plt.ylabel(figure_options.ylabel, labelpad=30)
# save_figure(figure_options, fig)
fig.savefig(next_name(env["pd-work"]), bbox_inches="tight")
plt.show()
def main(env, args):
# type: (Environment, argparse.Namespace) -> None
gil = GenomeInfoList.init_from_file(args.pf_genome_list)
pd_figures = os_join(env["pd-work"], "figures")
mkdir_p(pd_figures)
list_run_info = list()
for gi in tqdm(gil, total=len(gil)):
# get gms2 and toolp models
mod_gms2, mod_toolp = compare_gms2_and_toolp_motifs_for_gi(env, gi)
group = mod_gms2.items["GENOME_TYPE"].split("-")[1].upper()
mm_gms2 = MotifModel(mod_gms2.items["RBS_MAT"], None)
mm_toolp = MotifModel(mod_toolp.items["RBS_MAT"], None)
non_gms2 = GMS2Noncoding(mod_gms2.items["NON_MAT"])
df_gms2 = mm_gms2.pwm_to_df()
df_toolp = mm_toolp.pwm_to_df()
fig, axes = plt.subplots(1, 2, sharex="all", sharey="all", figsize=(8, 4))
# relative
rel_mat = lm.transform_matrix(df_gms2, from_type="probability", to_type="information")
lm.Logo(rel_mat, color_scheme="classic", ax=axes[0])
axes[0].set_ylim(*[0, 2])
axes[0].set_title("GeneMarkS-2")
# shannon
sha_mat = lm.transform_matrix(df_toolp, from_type="probability", to_type="information")
lm.Logo(sha_mat, color_scheme="classic", ax=axes[1])
axes[1].set_ylim(*[0, 2])
axes[1].set_title("StartLink+")
plt.tight_layout()
plt.savefig(next_name(pd_figures))
plt.show()
rel_gms2 = relative_entropy(mm_gms2, non_gms2)
rel_toolp = relative_entropy(mm_toolp, non_gms2)
gc = 100 * compute_gc_from_file(os_join(env["pd-data"], gi.name, "sequence.fasta"))
if not args.verified:
list_run_info.append({
"GC": gc,
"Accuracy": 100 - compare_gms2_start_predictions_with_motif_from_toolp(env, gi),
"RE GMS2": rel_gms2,
"RE toolp": rel_toolp
})
else:
# verified
comp = compare_gms2_start_predictions_with_motif_from_toolp_verified(env, gi, group=group)
list_run_info.append({
"Genome": fix_names(gi.name),
"Error": 100 - comp[0],
"Tool": "GMS2",
"RE": rel_gms2,
"GC": gc
})
list_run_info.append({
"Genome": fix_names(gi.name),
"Error": 100 - comp[1],
"Tool": "GMS2 with SL",
"RE": rel_toolp,
"GC": gc
})
print(list_run_info[-2:])
import sbsp_viz.sns as sns
if args.verified:
df = pd.DataFrame(list_run_info)
df.to_csv(next_name(env["pd-work"], ext="csv"))
sns.lineplot(df, "Genome", "Error", hue="Tool", figure_options=FigureOptions(
save_fig=next_name(env["pd-work"]),
xlabel="Genome",
ylabel="Error"))
sns.lineplot(df, "Genome", "RE", hue="Tool",
figure_options=FigureOptions(
save_fig=next_name(env["pd-work"]),
xlabel="Genome",
ylabel="Relative entropy",
))
else:
df = pd.DataFrame(list_run_info)
sns.scatterplot(df, "GC", "Accuracy",
figure_options=FigureOptions(
save_fig=next_name(env["pd-work"]),
xlabel="GC",
ylabel="Percentage of different 5' ends",
ylim=[0,10],
))
df.to_csv(next_name(env["pd-work"], ext="csv"))
sns.scatterplot(df, "RE GMS2", "RE toolp", identity=True, figure_options=FigureOptions(
save_fig=next_name(env["pd-work"])
))
print("Average Error: {}".format(df["Accuracy"].mean()))
df = pd.DataFrame(list_run_info)
df = df[df["Accuracy"] < 2].copy()
sns.scatterplot(df, "GC", "Accuracy",
figure_options=FigureOptions(
save_fig=next_name(env["pd-work"]),
xlabel="GC",
ylabel="Percentage of different 5' ends",
ylim=[0,10],
))
sns.scatterplot(df, "RE GMS2", "RE toolp", identity=True, figure_options=FigureOptions(
save_fig=next_name(env["pd-work"])
))
print("Average Error: {}".format(df["Accuracy"].mean()))
df.to_csv(next_name(env["pd-work"], ext="csv"))
def viz_summary_per_gcfid_per_step(env, df):
# type: (Environment, pd.DataFrame) -> None
# gather analysis for steps A, A+B, and A+B+C
list_df = list() # type: List[pd.DataFrame]
# compute total number of predictions per tool, per genome
for gcfid, df_group in df.groupby("GCFID", as_index=False):
df.loc[df_group.index, "Total SBSP"] = df.loc[df_group.index,
"SBSP"].sum()
df.loc[df_group.index, "Total GMS2"] = df.loc[df_group.index,
"GMS2"].sum()
df.loc[df_group.index, "Total GMS2=SBSP"] = df.loc[df_group.index,
"GMS2=SBSP"].sum()
# loop over steps A, A+B, and A+B+C and collect stats
tag = None
for step in ["A", "B", "C"]:
if tag is None:
tag = step
else:
tag += "+" + step
df_summary_per_gcfid = get_summary_per_gcfid(
df[df["Predicted-at-step"] <= step])
df_summary_per_gcfid["SBSP Step"] = tag
list_df.append(df_summary_per_gcfid)
df_per_gcfid_per_step = pd.concat(list_df, sort=False)
import matplotlib.pyplot as plt
# fig, ax = plt.subplots()
#
# sns.lineplot(df_per_gcfid_per_step, "SBSP Step", "SBSP", hue="GCFID", ax=ax,
# sns_kwargs={"palette": CM.get_map("verified")},
# legend=False
# )
# for l in ax.lines:
# l.set_linestyle("--")
#
# ax2 = ax.twinx()
# sns.lineplot(df_per_gcfid_per_step, "SBSP Step", "Sen(SBSP,NCBI)", hue="GCFID", ax=ax2,
# sns_kwargs={"palette": CM.get_map("verified")},)
#
# fo = FigureOptions(
# xlabel="SBSP Step",
# ylabel="Percentage",
# # ylim=[0, 105],
# save_fig=next_name(env["pd-work"])
# )
# FigureOptions.set_properties_for_axis(ax, fo)
# plt.subplots_adjust(bottom=0.2)
# handles, labels = ax.get_legend_handles_labels()
# ax.legend(handles=handles[1:], labels=labels[1:],
# loc="lower center", ncol=4, bbox_to_anchor=(0.5, -0.25))
#
# plt.savefig(fo.save_fig)
# plt.show()
fig, axes = plt.subplots(3, 2, sharex="all", sharey="row")
ax = axes[:, 0]
sns.lineplot(df_per_gcfid_per_step,
"SBSP Step",
"Sen(SBSP,NCBI)",
hue="GCFID",
ax=ax[0],
sns_kwargs={"palette": CM.get_map("verified")},
legend=False,
figure_options=FigureOptions(
ylabel="Error rate (\%)",
ylim=[0, 20],
))
sns.lineplot(df_per_gcfid_per_step,
"SBSP Step",
"Cov(SBSP,NCBI)",
hue="GCFID",
ax=ax[1],
sns_kwargs={"palette": CM.get_map("verified")},
legend=False,
figure_options=FigureOptions(ylabel="Percentage\nof Genes",
ylim=[0, None]))
sns.lineplot(df_per_gcfid_per_step,
"SBSP Step",
"SBSP",
hue="GCFID",
ax=ax[2],
sns_kwargs={"palette": CM.get_map("verified")},
legend=False,
figure_options=FigureOptions(ylabel="Number\nof Genes",
ylim=[0, None]))
fig.align_ylabels(ax)
# plt.savefig(next_name(env["pd-work"]))
# plt.show()
# fig, ax = plt.subplots(3, 1, sharex="all")
ax = axes[:, 1]
sns.lineplot(df_per_gcfid_per_step,
"SBSP Step",
"Sen(GMS2=SBSP,NCBI)",
hue="GCFID",
ax=ax[0],
sns_kwargs={"palette": CM.get_map("verified")},
legend=False,
figure_options=FigureOptions(
ylabel="Error",
ylim=[0, None],
))
sns.lineplot(df_per_gcfid_per_step,
"SBSP Step",
"Cov(GMS2=SBSP,NCBI)",
hue="GCFID",
ax=ax[1],
sns_kwargs={"palette": CM.get_map("verified")},
legend=False,
figure_options=FigureOptions(ylabel="Percentage of Genes",
ylim=[0, None]))
sns.lineplot(df_per_gcfid_per_step,
"SBSP Step",
"GMS2=SBSP",
hue="GCFID",
ax=ax[2],
sns_kwargs={"palette": CM.get_map("verified")},
figure_options=FigureOptions(ylabel="Number of Genes",
ylim=[0, None]))
ax[2].get_legend().remove()
fig.align_ylabels(ax)
for ax in axes.ravel():
ax.set_xlabel("Steps")
axes[0][0].set_title(TOOL)
axes[0][1].set_title(TOOLp)
fig.subplots_adjust(bottom=0.21)
# handles, labels = ax.get_legend_handles_labels()
# fig.legend(handles=handles[1:], labels=labels[1:], loc="lower center", ncol=4)#, bbox_to_anchor=(0.5, -0.25))
handles, labels = ax.get_legend_handles_labels()
labels[0] = "Genome"
fig.legend(handles=handles, labels=labels, loc="lower center",
ncol=3) #, bbox_to_anchor=(0.5, -0.25))
plt.savefig(next_name(env["pd-work"]))
plt.show()
# three plots
for gcfid, df_group in df.groupby("GCFID", as_index=False):
df.loc[df_group.index,
"Total SBSP"] = ((df_group["SBSP"]) & (df_group["NCBI"])).sum()
df.loc[df_group.index,
"Total GMS2"] = ((df_group["GMS2"]) & (df_group["NCBI"])).sum()
df.loc[df_group.index, "Total GMS2=SBSP"] = ((df_group["GMS2=SBSP"]) &
(df_group["NCBI"])).sum()
df_all = get_summary_per_gcfid(df)
# map column names for tables
columns = [
"GCFID", "NCBI", "Sen(SBSP,NCBI)", "Sen(GMS2,NCBI)",
"Sen(GMS2=SBSP,NCBI)", "Cov2(SBSP,NCBI)", "Cov2(GMS2,NCBI)",
"Cov2(GMS2=SBSP,NCBI)"
]
df_sen = df_all.copy()[columns].rename(columns={
"GCFID": "Genome",
"NCBI": "Verified",
"Sen(SBSP,NCBI)": "SBSP",
"Sen(GMS2,NCBI)": "GMS2",
"Sen(GMS2=SBSP,NCBI)": "GMS2=SBSP",
},
inplace=False)
df_sen[["Genome", "Verified", "SBSP", "GMS2",
"GMS2=SBSP"]].to_csv(os_join(env["pd-work"], "sensitivity.csv"),
index=False)
# print(df_all[["GCFID", "NCBI", "Cov2(SBSP,NCBI)", "Cov2(GMS2,NCBI)", "Cov2(GMS2=SBSP,NCBI)"]].to_string(index=False))
df_cov = df_all[columns].rename(columns={
"GCFID": "Genome",
"NCBI": "Verified",
"Cov2(SBSP,NCBI)": "SBSP",
"Cov2(GMS2,NCBI)": "GMS2",
"Cov2(GMS2=SBSP,NCBI)": "GMS2=SBSP",
},
inplace=False)
df_cov[["Genome", "Verified", "SBSP", "GMS2",
"GMS2=SBSP"]].to_csv(os_join(env["pd-work"], "coverage.csv"),
index=False)
def main(env, args):
# type: (Environment, argparse.Namespace) -> None
df = pd.read_csv(args.pf_data)
df["chunk-size"] /= 1000
import matplotlib.pyplot as plt
fig, ax = plt.subplots()
sns.lineplot(df[df["Tool"] == "SBSP"], "chunk-size", "percentage-common-3prime-and-5prime-from-common-3prime",
hue="Genome",
sns_kwargs={"palette": CM.get_map("verified"), "linestyle": "dashed"},
ax=ax,
legend=False,
figure_options=FigureOptions(
xlabel="Chunk size (mb)",
ylabel="Accuracy",
ylim=[74, 101],
save_fig=next_name(env["pd-work"])
))
for l in ax.lines:
l.set_linestyle("--")
sns.lineplot(df[df["Tool"] == "GMS2"], "chunk-size", "percentage-common-3prime-and-5prime-from-common-3prime",
hue="Genome",
sns_kwargs={"palette": CM.get_map("verified")},
legend_loc="best",
legend_ncol=2,
ax=ax)
if args.with_mgm:
y_max = ax.get_ylim()[1]
ax.axvline(50, 0, y_max, color="grey", linestyle="dashed")
ax.axhline(74, 5, 49, color="grey", linestyle="dashed")
ax.annotate("MGM", (5, 72))
if "MGM" in set(df["Tool"]):
sns.lineplot(df[df["Tool"] == "MGM"], "chunk-size", "percentage-common-3prime-and-5prime-from-common-3prime",
hue="Genome",
sns_kwargs={"palette": CM.get_map("verified"), "linestyle": "-."},
ax=ax,
legend=False)
for l in ax.lines[len(ax.lines)-5:]:
l.set_linestyle(":")
fo = FigureOptions(
xlabel="Chunk size (mb)",
ylabel="Accuracy",
ylim=[74,101],
save_fig=next_name(env["pd-work"])
)
FigureOptions.set_properties_for_axis(ax, fo)
plt.savefig(fo.save_fig)
plt.show()
def analyze_independent_predictions(max_candidates, sen_a, sen_b):
# type: (int, float, float) -> None
sensitivities = {
"Random": sensitivity_random,
"Independent": sensitivity_independent,
"Fully dependent": sensitivity_fully_dependent
}
agree_given_pred = {
"Random": agree_given_pred_random,
"Independent": agree_given_pred_independent,
"Fully dependent": agree_given_pred_fully_dependent
}
df = compute_data(sensitivities, agree_given_pred, max_candidates)
plot_sensitivities_vs_num_candidates(sensitivities, max_candidates, sen_a,
sen_b)
sns.lineplot(
df[(df["Sensitivity A"] == 0.9) & (df["Sensitivity B"] == 0.9)],
"Number of candidates",
"Probability",
hue="Condition",
sns_kwargs={"palette": CM.get_map("independence-conditions")},
legend_loc="best",
figure_options=FigureOptions(
save_fig=next_name("."),
ylabel=r"$P(y=s|x_1=y, x_2=y)$",
# xlim=[None, 40]
))
# error
df["1 - Probability"] = 1 - df["Probability"]
sns.lineplot(
df[(df["Sensitivity A"] == 0.9) & (df["Sensitivity B"] == 0.9)],
"Number of candidates",
"1 - Probability",
hue="Condition",
sns_kwargs={"palette": CM.get_map("independence-conditions")},
legend_loc="best",
figure_options=FigureOptions(
save_fig=next_name("."),
ylabel=r"$P(y\neq s|x_1=y, x_2=y)$",
# xlim=[None, 40]
))
import matplotlib.pyplot as plt
fig, axes = plt.subplots(1, 2, sharey="all", figsize=(10, 4))
sns.lineplot(df[(df["Sensitivity A"] == 0.9)
& (df["Sensitivity B"] == 0.9)],
"Number of candidates",
"Probability",
hue="Condition",
sns_kwargs={"palette": CM.get_map("independence-conditions")},
ax=axes[0],
legend=False,
figure_options=FigureOptions(title="Sensitivity = 0.9", ))
sns.lineplot(df[(df["Sensitivity A"] == df["Sensitivity B"])
& (df["Number of candidates"] == 25)],
"Sensitivity A",
"Probability",
hue="Condition",
ax=axes[1],
sns_kwargs={"palette": CM.get_map("independence-conditions")},
figure_options=FigureOptions(
ylim=[0, 1.05],
xlim=[0, 1],
xlabel="Sensitivity",
title="Number of candidates = 25",
))
save_figure(FigureOptions(save_fig=next_name(".")), fig)
plt.show()
df_tmp = df[(df["Sensitivity A"] == df["Sensitivity B"])
& (df["Condition"] == "Independent") &
(df["Sensitivity A"].isin(
{0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9}))]
df_tmp.rename(columns={"Sensitivity A": "Sensitivity"}, inplace=True)
sns.lineplot(
df_tmp,
"Number of candidates",
"Probability",
hue="Sensitivity",
figure_options=FigureOptions(
# ylim=[0, 1.05],
# xlim=[0, 1],
title="Independent algorithms",
save_fig=next_name(".")),
)
# for condition in set(df["Condition"]):
#
# sns.kdeplot(
# df[(df["Condition"] == condition) & (df["Sensitivity A"] == df["Sensitivity B"])],
# "Sensitivity A", "Number of candidates", "Probability",
# figure_options=FigureOptions(
# title=condition
# ))
import matplotlib.pyplot as plt
fig, axes = plt.subplots(1, 2, sharey="all", figsize=(10, 4))
sns.lineplot(df[(df["Sensitivity A"] == 0.9)
& (df["Sensitivity B"] == 0.9)],
"Number of candidates",
"Agree given prediction",
hue="Condition",
sns_kwargs={"palette": CM.get_map("independence-conditions")},
ax=axes[0],
legend=False,
figure_options=FigureOptions(title="Sensitivity = 0.9", ))
sns.lineplot(df[(df["Sensitivity A"] == df["Sensitivity B"])
& (df["Number of candidates"] == 25)],
"Sensitivity A",
"Agree given prediction",
hue="Condition",
ax=axes[1],
sns_kwargs={"palette": CM.get_map("independence-conditions")},
figure_options=FigureOptions(
ylim=[0, 1.05],
xlim=[0, 1],
xlabel="Sensitivity",
title="Number of targets = 25",
))
save_figure(FigureOptions(save_fig=next_name(".")), fig)
plt.show()