seqcol, "site_str_pos", "site_str_score", "site_str_ori",
"site_str_core", "site_wk_pos", "site_wk_score", "site_wk_ori",
"site_wk_core", "distance", "orientation"
]]
return df.merge(posdf, on=seqcol)
if __name__ == "__main__":
basepath = "/Users/vincentiusmartin/Research/chip2gcPBM/chip2probe"
pd.set_option("display.max_columns", None)
# using pwm
pwm_ets = PWM("%s/input/site_models/pwm/ets1.txt" % basepath, log=True)
kompas_ets = Kompas(
"/Users/vincentiusmartin/Research/chip2gcPBM/chip2probe/input/site_models/kompas/Ets1_kmer_alignment.txt",
core_start=11,
core_end=15,
core_center=12)
df = pd.read_csv("%s/output/homotypic/training/seqlbled.csv" %
basepath).drop_duplicates()
df = df[(df["label"] == "cooperative") | (df["label"] == "independent")]
train = get_sites_pos(df, kompas_ets, pwm_ets, seqcol="sequence")
train = train[(train["site_wk_score"] != -999)
& (train["site_str_score"] != -999)]
print(train["label"].value_counts())
print(train["site_str_core"].unique())
train["orientation"].replace({"-/+": "+/-"}, inplace=True)
train.to_csv("train_pwm.csv", index=False)
# using custom imads model
"/Users/faricazjj/Box/homotf/chip2probe/chip2probe/data/imads/runx1/Runx1_10nM_Bound_filtered_normalized_logistic_transformed_20bp_GCGGT_1a2a3mer_format.model",
"/Users/faricazjj/Box/homotf/chip2probe/chip2probe/data/imads/runx1/Runx1_10nM_Bound_filtered_normalized_logistic_transformed_20bp_GTGGC_1a2a3mer_format.model",
"/Users/faricazjj/Box/homotf/chip2probe/chip2probe/data/imads/runx1/Runx1_10nM_Bound_filtered_normalized_logistic_transformed_20bp_GTGGG_1a2a3mer_format.model",
"/Users/faricazjj/Box/homotf/chip2probe/chip2probe/data/imads/runx1/Runx1_10nM_Bound_filtered_normalized_logistic_transformed_20bp_GTGGT_1a2a3mer_format.model"
]
}
imads_cutoff = {'ets1': 0.2128, 'runx1': 0.3061}
escores = {}
models = {}
proteins = ['ets1', 'runx1']
for tf in proteins:
escores[tf] = PBMEscore(escore_long_paths[tf])
models[tf] = Kompas(protein=tf,
threshold=mutate_cutoff,
kmer_align_path=kmer_align_paths[tf])
es_preds = {}
esplots = {}
model_preds = {}
model_plots = {}
colors = [('crimson', 'plum'), ('steelblue', 'lightblue')]
# initialize escore and model objects for each protein
for protein in proteins:
protein_num = proteins.index(protein)
es_preds[protein] = escores[protein].predict_sequences(df,
key_colname="key")
esplots[protein] = escores[protein].make_plot_data(
es_preds[protein], color=colors[protein_num][0])
"Name", "Sequence", "label"
]]
train00["Sequence"] = train00["Sequence"].apply(
lambda x: bio.revcompstr(x))
train00 = get_sites_pos(train00, kompas, pwm)
train = pd.concat([train[train["orientation"] != "-/-"], train00])
return train.drop_duplicates()
if __name__ == "__main__":
pd.set_option("display.max_columns", None)
# using pwm
pwm_ets = PWM("input/sitemodels/ets1.txt", log=True)
kompas_ets = Kompas("input/sitemodels/Ets1_kmer_alignment.txt",
core_start=11,
core_end=15,
core_center=12)
df = pd.read_csv(
"output/Ets1Ets1/label_pr/ets_ets_seqlabeled.csv").drop_duplicates()
df = df[(df["label"] == "cooperative") | (df["label"] == "independent")]
train = gen_training(df, pwm_ets, kompas_ets)
print(train["label"].value_counts())
train.to_csv("output/Ets1Ets1/training/train_ets1_ets1.tsv",
index=False,
sep="\t")
train.rename(columns={
'site_str_score': 'Binding strength of the stronger site',
'site_wk_score': 'Binding strength of the weaker site'
},
#escore_pred_df = escore.predict_sequences(df, key_colname="key")
# ========= iMADS =========
# load imads model, the cores should be loaded in the same order with the model
# imads_paths = ["input/site_models/imads_models/Ets1_w12_GGAA.model", "input/site_models/imads_models/Ets1_w12_GGAT.model"]
# imads_cores = ["GGAA", "GGAT"]
# imads_models = [iMADSModel(path, core, 12, [1, 2, 3]) for path, core in zip(imads_paths, imads_cores)]
# imads = iMADS(imads_models, 0.19) # 0.2128
#
# imads_pred = imads.predict_sequence(single_sequence)
# imads_pred_list = imads.predict_sequences(many_sequences)
# imads_pred_df = imads.predict_sequences(df, key_colname="key")
# ========= Kompas =========
kompas = Kompas("input/site_models/kompas/Ets1_kmer_alignment.txt",
core_start=11,
core_end=15,
core_center=12)
kompas_pred = kompas.predict_sequence(single_sequence)
kompas_pred_list = kompas.predict_sequences(many_sequences)
# ========= Plot the sequence =========
# Make the plot objects, make_plot_data accepts prediction result
# pwm_plotr = pwmr.make_plot_data(pwm_pred_listr)
# pwm_plote = pwme.make_plot_data(pwm_pred_liste, color="green")
escore_plot = escore.make_plot_data(escore_pred_list, color="#ED2024")
# imads_plot = imads.make_plot_data(imads_pred_list)
kompas_plot = kompas.make_plot_data(kompas_pred_list, color="#F7A091")
# Generate sequence plot
sp = SitesPlotter()