app.py

import streamlit as st
from st_aggrid import GridOptionsBuilder, AgGrid, GridUpdateMode, DataReturnMode
import numpy as np
import pandas as pd
import sklearn
import xgboost
seed=42

data = pd.read_csv("annotations_dataset.csv")
data = data.set_index("Gene")

training_data = pd.read_csv("./selected_features_training_data.csv", header=0)
training_data.columns = [
    regex.sub("_", col) if any(x in str(col) for x in set(("[", "]", "<"))) else col
    for col in training_data.columns.values
]

training_data["BPlabel_encoded"] = training_data["BPlabel"].map(
    {"most likely": 1, "probable": 0.75, "least likely": 0.1}
)
Y = training_data["BPlabel_encoded"]
X = training_data.drop(columns=["BPlabel_encoded","BPlabel"])
xgb = xgboost.XGBRegressor(
    n_estimators=40,
    learning_rate=0.2,
    max_depth=4,
    reg_alpha=1,
    reg_lambda=1,
    random_state=seed,
    objective="reg:squarederror",
)


xgb.fit(X, Y)

predictions = list(xgb.predict(data))

output = pd.Series(data=predictions, index=data.index, name="XGB_Score")
df_total = pd.concat([data, output], axis=1)
df_total['XGB_Score'] = round(df_total['XGB_Score'], 2)

st.title('Blood Pressure Gene Prioritisation Post-GWAS')
st.markdown("""
A machine learning pipeline for predicting disease-causing genes post-genome-wide association study in blood pressure.
""")

gene_input = st.text_input('Input HGNC Gene') 
df = df_total[df_total.index == gene_input] 
st.dataframe(df)


st.markdown("""
Total Gene Prioritisation Results:
""")

st.dataframe(df_total)