import streamlit as st
import time
import concurrent.futures
import json
# import tensorflow
from gensim.models import Word2Vec
import pandas as pd
import threading
import matplotlib.pyplot as plt
import squarify
import numpy as np
import re
import urllib.request
import random
import plotly.express as px
import plotly.graph_objs as go
from streamlit.components.v1 import html
st.set_page_config(page_title="OncoDigger", page_icon=":microscope:", layout="wide", # centered
initial_sidebar_state="auto",
menu_items={'About': "OncoDigger is a Natural Language Processing (NLP) that harnesses Word2Vec to mine"
" insight from pubmed abstracts. Created by Jimmie E. Fata, PhD, fata4science@gmail.com"})
analytics_code = '''
'''
html(analytics_code, height=0)
# Define the HTML and CSS styles
st.markdown("""
""", unsafe_allow_html=True)
st.markdown("""
""", unsafe_allow_html=True)
st.header(":red[*O*]nco:red[*D*]igger")
st.subheader(
"A web app designed to explore massive amounts of :red[*PubMed abstracts*] for a deeper understanding of your research. Results are driven "
"by Machine Learning and Natural Language Processing algorithms, which allow you to scan and mine information from hundreds of thousands of abstracts in seconds.")
def custom_subheader(text, identifier, font_size):
st.markdown(f"{text}
", unsafe_allow_html=True)
custom_subheader("To begin, simply select a cancer corpus from the left sidebar and enter a keyword "
"you wish to explore within the corpus. OncoDigger will determine the top words, "
"genes, drugs, phytochemicals, and compounds that are contextually and semantically related "
"to your input, both directly and indirectly. Dive in and enjoy the exploration!",
"unique-id", 18)
st.markdown("---")
# Define the correct password
# CORRECT_PASSWORD = "123"
# Define a function to check if the password is correct
# def authenticate(password):
# if password == CORRECT_PASSWORD:
# return True
# else:
# return False
#
# # Create a Streamlit input field for the password
# password = st.text_input("Enter password:", type="password")
#
# # If the password is correct, show the app content
# if authenticate(password):
opt = st.sidebar.radio("Select a PubMed Corpus", options=('Breast Cancer corpus', 'Lung Cancer corpus',
'Skin Cancer corpus', 'Colorectal Cancer corpus',
'Prostate Cancer corpus', 'Lymphoma Cancer corpus'))
# if opt == "Clotting corpus":
# model_used = ("pubmed_model_clotting")
# num_abstracts = 45493
# database_name = "Clotting"
# if opt == "Neuroblastoma corpus":
# model_used = ("pubmed_model_neuroblastoma")
# num_abstracts = 29032
# database_name = "Neuroblastoma"
if opt == "Breast Cancer corpus":
model_used = ("breast_cancer_pubmed_model")
num_abstracts = 204381
database_name = "Breast_cancer"
if opt == "Lung Cancer corpus":
model_used = ("lung_cancer_pubmed_model")
num_abstracts = 143886
database_name = "Lung_cancer"
if opt == "Colorectal Cancer corpus":
model_used = ("colorectal_cancer_pubmed_model")
num_abstracts = 140000
database_name = "Colorectal_cancer"
if opt == "Prostate Cancer corpus":
model_used = ("prostate_cancer_pubmed_model")
num_abstracts = 89782
database_name = "Prostate_cancer"
if opt == "Skin Cancer corpus":
model_used = ("skin_cancer_pubmed_model")
num_abstracts = 176568
database_name = "Skin_cancer"
if opt == "Lymphoma corpus":
model_used = ("lymphoma_cancer_pubmed_model")
num_abstracts = 79000
database_name = "Lymphoma_cancer"
st.header(f":blue[{database_name} Pubmed corpus.]")
text_input_value = st.text_input(f"Enter one term to search within the {database_name} corpus")
query = text_input_value
query = query.lower()
query = query.strip() # This line will remove any leading or trailing spaces
query = re.sub("[,.?!&*;:]", "", query)
query = re.sub(" ", "-", query)
# matches = [" "]
# if any([x in query for x in matches]):
# st.write("Please only enter one term or a term without spaces")
# # query = input ("Enter your keyword(s):")
if query:
bar = st.progress(0)
time.sleep(.05)
st.caption(
f"Searching {num_abstracts} {database_name} PubMed primary abstracts covering 1990-2022 (Reviews not included)")
for i in range(10):
bar.progress((i + 1) * 10)
time.sleep(.1)
try:
model = Word2Vec.load(f"{model_used}") # you can continue training with the loaded model!
words = list(model.wv.key_to_index)
X = model.wv[model.wv.key_to_index]
# print(model.wv['bfgf'])
model2 = model.wv[query]
# print(model.wv.similar_by_word('bfgf', topn=50, restrict_vocab=None))
df = pd.DataFrame(X)
if 'melanin' in model.wv.key_to_index:
print("The term 'melanin' is present in the model.")
else:
print("The term 'melanin' is not present in the model.")
def get_compound_ids(compound_names):
with concurrent.futures.ThreadPoolExecutor() as executor:
compound_ids = list(executor.map(get_compound_id, compound_names))
return compound_ids
import requests
def get_compound_id(compound_name):
url = f"http://rest.kegg.jp/find/compound/{compound_name}"
response = requests.get(url)
if response.status_code == 200:
result = response.text.split('\n')
if result[0]:
compound_id = result[0].split('\t')[0]
return compound_id
return None
# except:
# st.error("Term occurrence is too low - please try another term")
# st.stop()
st.markdown("---")
try:
table = model.wv.most_similar_cosmul(query, topn=10000)
table = (pd.DataFrame(table))
table.index.name = 'Rank'
table.columns = ['Word', 'SIMILARITY']
pd.set_option('display.max_rows', None)
table2 = table.copy()
st.markdown(
f""
f"Top 10000 words in an interactive embedding map most similar to {query} in {database_name} "
f"PubMed corpus: Zoom in to the black diamond to find {query}
",
unsafe_allow_html=True)
# Set the max number of words to display
value_word = min(100, len(table2))
try:
value_word = min(50, len(table2))
# Get the top 50 similar words to the query
top_words = model.wv.most_similar_cosmul(query, topn=10000)
words = [word for word, sim in top_words]
words = [word.replace(' ', '-') for word in words]
sims = [sim for word, sim in top_words]
X_top = model.wv[words]
# Compute similarities between query and top 100 words
sims_query_top = sims # print(sims_query_top)
except Exception as e:
print("Error:", e)
# Generate a 2D scatter plot of word embeddings using Plotly
fig = px.scatter(x=X_top[:, 0], y=X_top[:, 1], color=sims_query_top, color_continuous_scale="RdYlGn", )
# Change background color to black
fig.update_layout(plot_bgcolor='#CCFFFF')
# Change color of text to white
fig.update_layout(xaxis=dict(gridcolor='#CCFFFF', color='blue'),
yaxis=dict(gridcolor='#CCFFFF', color='blue'))
# fig.update_traces(hovertemplate='%{hovertext}
Similarity score: %{customdata[0]:.2f}')
# fig.update_layout(title=dict(
# text=f"Top 10000 words in an interactive embedding map for {query} in {database_name} PubMed corpus"
# f": Zoom in to the black diamond to find {query}", x=0.5, y=1, xanchor='center', yanchor='top',
# font=dict(color='black')))
fig.update_coloraxes(colorbar_title=f"Similarity with {query}")
# Represent query as a large red diamond
fig.add_trace(go.Scatter(x=[model.wv[query][0]], y=[model.wv[query][1]], mode='markers',
marker=dict(size=12, color='black', symbol='diamond'), name=query, hovertext=query,
showlegend=False))
# Add label for the query above the diamond
fig.add_trace(go.Scatter(x=[model.wv[query][0]], y=[model.wv[query][1]], mode='text', text=[query],
textposition='top right', textfont=dict(color='blue', size=10), hoverinfo='none',
showlegend=False))
# Add circles for the top 50 similar words
fig.add_trace(go.Scatter(x=X_top[:, 0], y=X_top[:, 1], mode='markers',
marker=dict(size=2, color=sims_query_top, colorscale='RdYlGn', symbol='circle'),
text=words, customdata=sims, name=''))
fig.update(layout_coloraxis_showscale=True)
fig.update_layout(autosize=True, paper_bgcolor="#CCFFFF", margin=dict(t=0, b=0, l=0, r=0))
fig.update_annotations(visible=False)
st.plotly_chart(fig, use_container_width=True)
st.markdown(
f"Top {value_word} "
f"words contextually and semantically similar to "
f"{query} within the {database_name} corpus. "
f"Click on the squares to expand and also the PubMed and Wikipedia links for more word information
",
unsafe_allow_html=True)
short_table = table2.head(value_word).round(2)
short_table.index += 1
short_table.index = (1 / short_table.index) * 10
sizes = short_table.index.tolist()
short_table.set_index('Word', inplace=True)
table2["SIMILARITY"] = 'Similarity Score ' + table2.head(value_word)["SIMILARITY"].round(2).astype(str)
rank_num = list(short_table.index.tolist())
df = short_table
df['text'] = short_table.index
df['href'] = [f'https://pubmed.ncbi.nlm.nih.gov/?term={database_name}%5Bmh%5D+NOT+review%5Bpt%5D' \
'+AND+english%5Bla%5D+AND+hasabstract+AND+1990:2022%5Bdp%5D+AND+' + c for c in short_table.index]
df['href2'] = [f'https://en.wikipedia.org/wiki/' + c for c in short_table.index]
df.loc[:, 'database'] = database_name
fig = px.treemap(df, path=[short_table.index], values=sizes, custom_data=['href', 'text', 'database', 'href2'],
hover_name=(table2.head(value_word)['SIMILARITY']))
fig.update(layout_coloraxis_showscale=False)
fig.update_layout(autosize=True, paper_bgcolor="#CCFFFF", margin=dict(t=0, b=0, l=0, r=0))
fig.update_annotations(visible=False)
fig.update_traces(marker=dict(cornerradius=5), root_color="#CCFFFF", hovertemplate=None,
hoverlabel_bgcolor="lightblue", hoverlabel_bordercolor="#000000", texttemplate="
%{customdata[1]}
"
"PubMed"
"
Wikipedia"
"")
fig.update_layout(uniformtext=dict(minsize=15), treemapcolorway=["lightgreen"])
# st.pyplot(fig2)
st.plotly_chart(fig, use_container_width=True)
# st.caption(
# "Gene designation and database provided by HUGO Gene Nomenclature Committee (HGNC): https://www.genenames.org/")
# st.caption("Gene designation add in exceptions [p21, p53, her2, her3]")
csv = table2.head(value_word).to_csv().encode('utf-8')
st.download_button(label=f"download top {value_word} words (csv)", data=csv,
file_name=f'{database_name}_words.csv', mime='text/csv')
except:
st.warning(
f"This selection exceeds the number of similar words related to {query} within the {database_name} corpus, please choose a lower number")
except KeyError:
st.warning(
"This word is not found in the corpus, it could be because it is not spelled correctly or could be that it does not have enough representation within the corpus, please try again")
# try:
# value_word = min(50, len(table2))
# # Get the top 50 similar words to the query
# top_words = model.wv.most_similar_cosmul(query, topn=value_word)
# words = [word for word, sim in top_words]
# words = [word.replace(' ', '-') for word in words]
# sims = [sim for word, sim in top_words]
# X_top = model.wv[words]
#
# # Compute similarities between query and top 100 words
# sims_query_top = sims # print(sims_query_top)
# except Exception as e:
# print("Error:", e)
#
#
# # Generate a 3D scatter plot of word embeddings using Plotly
# fig = px.scatter_3d(x=X_top[:, 0], y=X_top[:, 1], z=X_top[:, 2], text=words, color=sims_query_top,
# color_continuous_scale="RdYlGn", hover_name=words, hover_data={"color": sims_query_top})
#
# # Change background color to black
# fig.update_layout(scene=dict(bgcolor='#CCFFFF'))
#
# # Change color of text to white
# fig.update_layout(scene=dict(xaxis=dict(backgroundcolor='#CCFFFF', color='blue'),
# yaxis=dict(backgroundcolor='#CCFFFF', color='blue'),
# zaxis=dict(backgroundcolor='#CCFFFF', color='blue')))
#
# fig.update_traces(hovertemplate='%{hovertext}
Similarity score: %{customdata[0]:.2f}')
# fig.update_layout(title=dict(text=f"Word embedding map for {query} in {database_name} PubMed corpus", x=0.5, y=0.95,
# xanchor='center', yanchor='top', font=dict(color='black')),
# scene=dict(xaxis_title="Dimension 1", yaxis_title="Dimension 2", zaxis_title="Dimension 3"))
# fig.update_coloraxes(colorbar_title="Similarity with query")
#
# # Represent query as a large red diamond
# fig.add_trace(go.Scatter3d(x=[model.wv[query][0]], y=[model.wv[query][1]], z=[model.wv[query][2]], mode='markers',
# marker=dict(size=7, color='black', symbol='diamond'), name=query, hovertext=query, showlegend=False))
#
# # Add label for the query above the diamond
# fig.add_trace(
# go.Scatter3d(x=[model.wv[query][0]], y=[model.wv[query][1]], z=[model.wv[query][2]], mode='text', text=[query],
# textposition='bottom center', textfont=dict(color='blue', size=10), hoverinfo='none', showlegend=False))
#
# # Add circles for the top 50 similar words
# fig.add_trace(go.Scatter3d(x=X_top[:, 0], y=X_top[:, 1], z=X_top[:, 2], mode='markers',
# marker=dict(size=2, color=sims_query_top, colorscale='RdYlGn', symbol='circle'),
# hovertemplate='%{text}
Similarity score: %{customdata[0]:.2f}',
# text=words, customdata=sims, name=''))
#
# fig.update(layout_coloraxis_showscale=True)
# fig.update_layout(autosize=True, paper_bgcolor="#CCFFFF", margin=dict(t=0, b=0, l=0, r=0))
# fig.update_annotations(visible=False)
#
# st.plotly_chart(fig, use_container_width=True)
st.markdown("---")
try:
df1 = table.copy()
df2 = pd.read_csv('Human Genes.csv')
m = df1.Word.isin(df2.symbol)
df1 = df1[m]
df1.rename(columns={'Word': 'Genes'}, inplace=True)
df_len = len(df1)
# print(len(df1))
# st.markdown(f"Populate a treemap to visualize "
# f"proteins contextually "
# f"and semantically similar to {query} "
# f"within the {database_name} corpus.
",
# unsafe_allow_html=True)
# Set the number of proteins to display
value_gene = min(df_len, 100)
st.markdown(
f"Top {value_gene} "
f"human genes contextually and semantically similar to "
f"{query} within the {database_name} corpus. Click on the squares to expand and also the Pubmed and GeneCard links for more gene information
",
unsafe_allow_html=True)
df11 = df1.head(value_gene).copy()
df11.index = (1 / df11.index) * 10000
sizes = df11.index.tolist()
df11.set_index('Genes', inplace=True)
df4 = df1.copy()
# print(df4.head(10))
df4["SIMILARITY"] = 'Similarity Score ' + df4.head(value_gene)["SIMILARITY"].round(2).astype(str)
df4.reset_index(inplace=True)
# df4 = df4.rename(columns={'Protein': 'symbol2'})
# print(df4)
# # Use df.query to get a subset of df1 based on ids in df2
# subset = df4.head(value_gene).query('symbol2 in @df2b.symbol2')
# # Use merge to join the two DataFrames on id
# result = pd.merge(subset, df2b, on='symbol2')
# print(result)
if value_gene <= df_len:
# Define the `text` column for labels and `href` column for links
df11['text'] = df11.index
df11['href'] = [f'https://pubmed.ncbi.nlm.nih.gov/?term={database_name}%5Bmh%5D+NOT+review%5Bpt%5D' \
'+AND+english%5Bla%5D+AND+hasabstract+AND+1990:2022%5Bdp%5D+AND+' + c for c in df11['text']]
df11['href2'] = [f'https://www.genecards.org/cgi-bin/carddisp.pl?gene=' + c for c in df11['text']]
assert isinstance(df11, object)
df11['database'] = database_name
# df11['name'] = [c for c in result['Approved name']]
# Create the treemap using `px.treemap`
fig = px.treemap(df11, path=[df11['text']], values=sizes, custom_data=['href', 'database', 'href2', 'text'],
hover_name=(df4.head(value_gene)['SIMILARITY']))
fig.update(layout_coloraxis_showscale=False)
fig.update_layout(autosize=True, paper_bgcolor="#CCFFFF", margin=dict(t=0, b=0, l=0, r=0))
fig.update_annotations(visible=False)
fig.update_traces(marker=dict(cornerradius=5), root_color="#CCFFFF", hovertemplate=None,
hoverlabel_bgcolor="lightblue", hoverlabel_bordercolor="#000000",
texttemplate="%{customdata[3]}
"
"PubMed"
"
GeneCard"
"")
fig.update_layout(uniformtext=dict(minsize=15), treemapcolorway=["LightPink"])
# # display the treemap in Streamlit
# with treemap2:
# st.pyplot(fig2)
st.plotly_chart(fig, use_container_width=True)
# st.caption(
# "Gene designation and database provided by KEGG homo sapien gene list: https://rest.kegg.jp/list/hsa")
# st.caption("Gene information provided by GeneCards: https://www.genecards.org//")
st.caption(
"Human gene designation and database provided by HUGO Gene Nomenclature Committee (HGNC): https://www.genenames.org/")
st.caption("Gene designation add in exceptions [p21, p53, her2, her3]")
st.caption("Gene information provided by GeneCards: https://www.genecards.org//")
csv = df1.head(value_gene).to_csv().encode('utf-8')
st.download_button(label=f"download top {value_gene} genes (csv)", data=csv,
file_name=f'{database_name}_genes.csv', mime='text/csv')
else:
st.warning(
f"This selection exceeds the number of similar proteins related to {query} within the {database_name} corpus, please choose a lower number")
value_gene = min(df_len, 50)
st.markdown(
f"3D interactive "
f"gene embedding map for {value_gene} genes most similar "
f"with {query} in {database_name} PubMed corpus
",
unsafe_allow_html=True)
try:
# Get the top 50 similar genes to the query
value_gene = min(df_len, 50)
top_words = model.wv.most_similar_cosmul(query, topn=value_gene)
words = df11.head(value_gene).index
words = [word.replace(' ', '-') for word in words]
# print(words)
sims = df4.head(value_gene)["SIMILARITY"].tolist()
# print(sims)
X_top = model.wv[words] # print(X_top)
except Exception as e:
print("Error:", e)
# Remove the text "Similarity Score" from each element in the sims list
sims_query_top = [float(sim.split()[-1]) for sim in sims]
# print(sims_query_top)
# Generate a 3D scatter plot of word embeddings using Plotly
fig2 = px.scatter_3d(x=X_top[:, 0], y=X_top[:, 1], z=X_top[:, 2], text=words, color=sims_query_top,
color_continuous_scale="RdYlGn", hover_name=words,
hover_data={"color": sims_query_top})
# Change background color to black
fig2.update_layout(scene=dict(bgcolor='#CCFFFF'))
# Change color of text to white
fig2.update_layout(scene=dict(xaxis=dict(backgroundcolor='#CCFFFF', color='blue'),
yaxis=dict(backgroundcolor='#CCFFFF', color='blue'),
zaxis=dict(backgroundcolor='#CCFFFF', color='blue')))
fig2.update_traces(
hovertemplate='%{hovertext}
Similarity score: %{customdata[0]:.2f}')
fig2.update_layout(
title=dict(text=f"", x=0.5, y=0.95,
xanchor='center', yanchor='top', font=dict(color='black')),
scene=dict(xaxis_title="Dimension 1", yaxis_title="Dimension 2", zaxis_title="Dimension 3"))
fig2.update_coloraxes(colorbar_title=f"Similarity with {query}")
# Represent query as a large red diamond
fig2.add_trace(
go.Scatter3d(x=[model.wv[query][0]], y=[model.wv[query][1]], z=[model.wv[query][2]], mode='markers',
marker=dict(size=7, color='black', symbol='diamond'), name=query, hovertext=query,
showlegend=False))
# Add label for the query above the diamond
fig2.add_trace(
go.Scatter3d(x=[model.wv[query][0]], y=[model.wv[query][1]], z=[model.wv[query][2]], mode='text',
text=[query], textposition='bottom center', textfont=dict(color='blue', size=10),
hoverinfo='none', showlegend=False))
# Add circles for the top 50 similar words
fig2.add_trace(go.Scatter3d(x=X_top[:, 0], y=X_top[:, 1], z=X_top[:, 2], mode='markers',
marker=dict(size=2, color=sims_query_top, colorscale='RdYlGn', symbol='circle'),
hovertemplate='%{text}
Similarity score: %{customdata[0]:.2f}',
text=words, customdata=sims, name=''))
fig2.update(layout_coloraxis_showscale=True)
fig2.update_layout(autosize=True, paper_bgcolor="#CCFFFF", margin=dict(t=0, b=0, l=0, r=0))
fig2.update_annotations(visible=False)
st.plotly_chart(fig2, use_container_width=True)
st.markdown("---")
# print()
# print("Human genes similar to " + str(query))
df1 = table.copy()
df2 = pd.read_csv('kegg_drug_list_lowercase.csv')
m = df1.Word.isin(df2.drugs)
df1 = df1[m]
df1.rename(columns={'Word': 'Drugs'}, inplace=True)
df_len = len(df1)
# print(len(df1))
# df1["Human Gene"] = df1["Human Gene"].str.upper()
# print(df1.head(50))
# print()
# df1.head(50).to_csv("clotting_sim2.csv", index=True, header=False)
# time.sleep(2)
# Create the slider with increments of 5 up to 100
# Remove the slider and set the value_compound to the minimum of the number of rows in the dataframe and 100
value_drug = min(df1.shape[0], 100)
# print(value_drug)
# st.markdown(f"Visualize "
# f"KEGG compounds contextually "
# f"and semantically similar to {query} "
# f"within the {database_name} corpus.
",
# unsafe_allow_html=True)
st.markdown(
f"Top {value_drug} "
f"Drugs contextually and semantically similar to "
f"{query} within the {database_name} corpus. Click on the squares to expand and the Pubmed and Wikipedia links for more compound information
",
unsafe_allow_html=True)
df13 = df1.head(value_drug).copy()
df13.index = (1 / df13.index) * 10000
sizes = df13.index.tolist()
df13.set_index('Drugs', inplace=True)
df6 = df1.copy()
# print(df4.head(10))
df6["SIMILARITY"] = 'Similarity Score ' + df6.head(value_drug)["SIMILARITY"].round(2).astype(str)
df6.reset_index(inplace=True)
# df4 = df4.rename(columns={'Protein': 'symbol2'})
# print(df4)
# # Use df.query to get a subset of df1 based on ids in df2
# subset = df4.head(value_gene).query('symbol2 in @df2b.symbol2')
# # Use merge to join the two DataFrames on id
# result = pd.merge(subset, df2b, on='symbol2')
# print(result)
if value_drug <= df_len:
# Define the `text` column for labels and `href` column for links
# Reset the index
df13.reset_index(inplace=True)
# Replace hyphens with spaces in the 'text' column
df13['Drugs'] = df13['Drugs'].str.replace('-', ' ')
# Set the 'text' column back as the index
df13.set_index('Drugs', inplace=True)
df13['text'] = df13.index
df13['href'] = [f'https://pubmed.ncbi.nlm.nih.gov/?term={database_name}%5Bmh%5D+NOT+review%5Bpt%5D' \
'+AND+english%5Bla%5D+AND+hasabstract+AND+1990:2022%5Bdp%5D+AND+' + c for c in df13['text']]
df13['href2'] = [f'https://en.wikipedia.org/wiki/' + c for c in df13['text']]
assert isinstance(df13, object)
df13['database'] = database_name
# df11['name'] = [c for c in result['Approved name']]
# Create the treemap using `px.treemap`
fig = px.treemap(df13, path=[df13['text']], values=sizes, custom_data=['href', 'database', 'href2', 'text'],
hover_name=(df6.head(value_drug)['SIMILARITY']))
fig.update(layout_coloraxis_showscale=False)
fig.update_layout(autosize=True, paper_bgcolor="#CCFFFF", margin=dict(t=0, b=0, l=0, r=0))
fig.update_annotations(visible=False)
fig.update_traces(marker=dict(cornerradius=5), root_color="#CCFFFF", hovertemplate=None,
hoverlabel_bgcolor="lightblue", hoverlabel_bordercolor="#000000",
texttemplate="%{customdata[3]}
"
"PubMed"
"
Wikipedia"
"")
fig.update_layout(uniformtext=dict(minsize=15), treemapcolorway=["Thistle"])
# # display the treemap in Streamlit
# with treemap2:
# st.pyplot(fig2)
st.plotly_chart(fig, use_container_width=True)
st.caption("Drug designation and database provided by KEGG: https://www.kegg.jp/kegg/drug/")
csv = df1.head(value_drug).to_csv().encode('utf-8')
st.download_button(label=f"download top {value_drug} drugs (csv)", data=csv,
file_name=f'{database_name}_drugs.csv', mime='text/csv')
else:
st.warning(
f"This selection exceeds the number of similar drugs related to {query} within the {database_name} corpus, please choose a lower number")
# try:
# value_drug = min(df_len, 50)
# top_words = model.wv.most_similar_cosmul(query, topn=value_drug)
# # print(top_words)
# words = df13.head(value_drug).index
# words = [word.replace(' ', '-') for word in words]
# # print(words)
# sims = df6.head(value_drug)["SIMILARITY"].tolist()
# # print(sims)
# X_top = model.wv[words]
# except Exception as e:
# print("Error:", e)
#
#
# # Remove the text "Similarity Score" from each element in the sims list
# sims_query_top = [float(sim.split()[-1]) for sim in sims]
# # print(sims_query_top)
#
# # Generate a 3D scatter plot of word embeddings using Plotly
# fig4 = px.scatter_3d(x=X_top[:, 0], y=X_top[:, 1], z=X_top[:, 2], text=words, color=sims_query_top,
# color_continuous_scale="RdYlGn", hover_name=words, hover_data={"color": sims_query_top})
#
# # Change background color to black
# fig4.update_layout(scene=dict(bgcolor='#CCFFFF'))
#
# # Change color of text to white
# fig4.update_layout(scene=dict(xaxis=dict(backgroundcolor='#CCFFFF', color='blue'),
# yaxis=dict(backgroundcolor='#CCFFFF', color='blue'),
# zaxis=dict(backgroundcolor='#CCFFFF', color='blue')))
#
# fig4.update_traces(hovertemplate='%{hovertext}
Similarity score: %{customdata[0]:.2f}')
# fig4.update_layout(
# title=dict(text=f"Word embedding map for {query} in {database_name} PubMed corpus", x=0.5, y=0.95,
# xanchor='center', yanchor='top', font=dict(color='black')),
# scene=dict(xaxis_title="Dimension 1", yaxis_title="Dimension 2", zaxis_title="Dimension 3"))
# fig4.update_coloraxes(colorbar_title="Similarity with query")
#
# # Represent query as a large red diamond
# fig4.add_trace(
# go.Scatter3d(x=[model.wv[query][0]], y=[model.wv[query][1]], z=[model.wv[query][2]], mode='markers',
# marker=dict(size=7, color='black', symbol='diamond'), name=query, hovertext=query,
# showlegend=False))
#
# # Add label for the query above the diamond
# fig4.add_trace(go.Scatter3d(x=[model.wv[query][0]], y=[model.wv[query][1]], z=[model.wv[query][2]], mode='text',
# text=[query], textposition='bottom center', textfont=dict(color='blue', size=10),
# hoverinfo='none', showlegend=False))
#
# # Add circles for the top 50 similar words
# fig4.add_trace(go.Scatter3d(x=X_top[:, 0], y=X_top[:, 1], z=X_top[:, 2], mode='markers',
# marker=dict(size=2, color=sims_query_top, colorscale='RdYlGn', symbol='circle'),
# hovertemplate='%{text}
Similarity score: %{customdata[0]:.2f}',
# text=words, customdata=sims, name=''))
#
# fig4.update(layout_coloraxis_showscale=True)
# fig4.update_layout(autosize=True, paper_bgcolor="#CCFFFF", margin=dict(t=0, b=0, l=0, r=0))
# fig4.update_annotations(visible=False)
#
# st.plotly_chart(fig4, use_container_width=True)
# st.markdown("---")
st.markdown("---")
# print()
# print("Human genes similar to " + str(query))
df1 = table.copy()
df2 = pd.read_csv('phytochemicals.csv')
m = df1.Word.isin(df2.phyto)
df1 = df1[m]
df1.rename(columns={'Word': 'Phytochemical'}, inplace=True)
df_len = len(df1)
# print(len(df1))
# df1["Human Gene"] = df1["Human Gene"].str.upper()
# print(df1.head(50))
# print()
# df1.head(50).to_csv("clotting_sim2.csv", index=True, header=False)
# time.sleep(2)
# Create the slider with increments of 5 up to 100
# Remove the slider and set the value_compound to the minimum of the number of rows in the dataframe and 100
value_phyto = min(df1.shape[0], 100)
# st.markdown(f"Visualize "
# f"KEGG compounds contextually "
# f"and semantically similar to {query} "
# f"within the {database_name} corpus.
",
# unsafe_allow_html=True)
st.markdown(
f"Top {value_phyto} "
f"Phytochemicals contextually and semantically similar to "
f"{query} within the {database_name} corpus. "
f"Click on the squares to expand and also the Pubmed and Wikipedia links for more compound information
",
unsafe_allow_html=True)
df15 = df1.head(value_phyto).copy()
df15.index = (1 / df15.index) * 10000
sizes = df15.index.tolist()
df15.set_index('Phytochemical', inplace=True)
df8 = df1.copy()
# print(df4.head(10))
df8["SIMILARITY"] = 'Similarity Score ' + df8.head(value_phyto)["SIMILARITY"].round(2).astype(str)
df8.reset_index(inplace=True)
# df4 = df4.rename(columns={'Protein': 'symbol2'})
# print(df4)
# # Use df.query to get a subset of df1 based on ids in df2
# subset = df4.head(value_gene).query('symbol2 in @df2b.symbol2')
# # Use merge to join the two DataFrames on id
# result = pd.merge(subset, df2b, on='symbol2')
# print(result)
if value_phyto <= df_len:
# Define the `text` column for labels and `href` column for links
# Reset the index
df15.reset_index(inplace=True)
# Replace hyphens with spaces in the 'text' column
df15['Phytochemical'] = df15['Phytochemical'].str.replace('-', ' ')
# Set the 'text' column back as the index
df15.set_index('Phytochemical', inplace=True)
df15['text'] = df15.index
df15['href'] = [f'https://pubmed.ncbi.nlm.nih.gov/?term={database_name}%5Bmh%5D+NOT+review%5Bpt%5D' \
'+AND+english%5Bla%5D+AND+hasabstract+AND+1990:2022%5Bdp%5D+AND+' + c for c in df15['text']]
df15['href2'] = [f'https://en.wikipedia.org/wiki/' + c for c in df15['text']]
assert isinstance(df15, object)
df15['database'] = database_name
# df11['name'] = [c for c in result['Approved name']]
# Create the treemap using `px.treemap`
fig = px.treemap(df15, path=[df15['text']], values=sizes, custom_data=['href', 'database', 'href2', 'text'],
hover_name=(df8.head(value_phyto)['SIMILARITY']))
fig.update(layout_coloraxis_showscale=False)
fig.update_layout(autosize=True, paper_bgcolor="#CCFFFF", margin=dict(t=0, b=0, l=0, r=0))
fig.update_annotations(visible=False)
fig.update_traces(marker=dict(cornerradius=5), root_color="#CCFFFF", hovertemplate=None,
hoverlabel_bgcolor="lightblue", hoverlabel_bordercolor="#000000",
texttemplate="%{customdata[3]}
"
"PubMed"
"
Wikipedia"
"")
fig.update_layout(uniformtext=dict(minsize=15), treemapcolorway=["LightSeaGreen"])
# # display the treemap in Streamlit
# with treemap2:
# st.pyplot(fig2)
st.plotly_chart(fig, use_container_width=True)
st.caption("Phytochemical designation and database provided by PhytoHub: https://phytohub.eu/")
csv = df1.head(value_phyto).to_csv().encode('utf-8')
st.download_button(label=f"download top {value_phyto} phytochemicals (csv)", data=csv,
file_name=f'{database_name}_phytochemicals.csv', mime='text/csv')
else:
st.warning(
f"This selection exceeds the number of similar pythochemicals related to {query} within the {database_name} corpus, please choose a lower number")
# try:
# value_phyto = min(df_len, 50)
# top_words = model.wv.most_similar_cosmul(query, topn=value_phyto)
# words = df15.head(value_phyto).index
# words = [word.replace(' ', '-') for word in words]
# # print(words)
# sims = df8.head(value_phyto)["SIMILARITY"].tolist()
# # print(sims)
# X_top = model.wv[words] # print(X_top)
# except Exception as e:
# print("Error:", e)
#
# # Remove the text "Similarity Score" from each element in the sims list
# sims_query_top = [float(sim.split()[-1]) for sim in sims]
# # print(sims_query_top)
#
# # Generate a 3D scatter plot of word embeddings using Plotly
# fig4 = px.scatter_3d(x=X_top[:, 0], y=X_top[:, 1], z=X_top[:, 2], text=words, color=sims_query_top,
# color_continuous_scale="RdYlGn", hover_name=words, hover_data={"color": sims_query_top})
#
# # Change background color to black
# fig4.update_layout(scene=dict(bgcolor='#CCFFFF'))
#
# # Change color of text to white
# fig4.update_layout(scene=dict(xaxis=dict(backgroundcolor='#CCFFFF', color='blue'),
# yaxis=dict(backgroundcolor='#CCFFFF', color='blue'),
# zaxis=dict(backgroundcolor='#CCFFFF', color='blue')))
#
# fig4.update_traces(hovertemplate='%{hovertext}
Similarity score: %{customdata[0]:.2f}')
# fig4.update_layout(
# title=dict(text=f"Word embedding map for {query} in {database_name} PubMed corpus", x=0.5, y=0.95,
# xanchor='center', yanchor='top', font=dict(color='black')),
# scene=dict(xaxis_title="Dimension 1", yaxis_title="Dimension 2", zaxis_title="Dimension 3"))
# fig4.update_coloraxes(colorbar_title="Similarity with query")
#
# # Represent query as a large red diamond
# fig4.add_trace(
# go.Scatter3d(x=[model.wv[query][0]], y=[model.wv[query][1]], z=[model.wv[query][2]], mode='markers',
# marker=dict(size=7, color='black', symbol='diamond'), name=query, hovertext=query,
# showlegend=False))
#
# # Add label for the query above the diamond
# fig4.add_trace(go.Scatter3d(x=[model.wv[query][0]], y=[model.wv[query][1]], z=[model.wv[query][2]], mode='text',
# text=[query], textposition='bottom center', textfont=dict(color='blue', size=10),
# hoverinfo='none', showlegend=False))
#
# # Add circles for the top 50 similar words
# fig4.add_trace(go.Scatter3d(x=X_top[:, 0], y=X_top[:, 1], z=X_top[:, 2], mode='markers',
# marker=dict(size=2, color=sims_query_top, colorscale='RdYlGn', symbol='circle'),
# hovertemplate='%{text}
Similarity score: %{customdata[0]:.2f}',
# text=words, customdata=sims, name=''))
#
# fig4.update(layout_coloraxis_showscale=True)
# fig4.update_layout(autosize=True, paper_bgcolor="#CCFFFF", margin=dict(t=0, b=0, l=0, r=0))
# fig4.update_annotations(visible=False)
#
# st.plotly_chart(fig4, use_container_width=True)
# st.markdown("---")
st.markdown("---")
# print()
# print("Human genes similar to " + str(query))
df1 = table.copy()
df2 = pd.read_csv('kegg_compounds_lowercase.csv')
m = df1.Word.isin(df2.compound)
df1 = df1[m]
df1.rename(columns={'Word': 'Compounds'}, inplace=True)
df_len = len(df1)
# df1["Human Gene"] = df1["Human Gene"].str.upper()
# print(df1.head(50))
# print()
# df1.head(50).to_csv("clotting_sim2.csv", index=True, header=False)
# time.sleep(2)
# Create the slider with increments of 5 up to 100
# Remove the slider and set the value_compound to the minimum of the number of rows in the dataframe and 100
value_compound = min(df1.shape[0], 100)
# st.markdown(f"Visualize "
# f"KEGG compounds contextually "
# f"and semantically similar to {query} "
# f"within the {database_name} corpus.
",
# unsafe_allow_html=True)
st.markdown(
f"Top {value_compound} "
f"Compounds contextually and semantically similar to "
f"{query} within the {database_name} corpus. "
f"Click on the squares to expand and the Pubmed, Wikipedia, and KEGG links for more compound information (may take time to load)
",
unsafe_allow_html=True)
df12 = df1.head(value_compound).copy()
df12.index = (1 / df12.index) * 10000
sizes = df12.index.tolist()
df12.set_index('Compounds', inplace=True)
df5 = df1.copy()
# print(df4.head(10))
df5["SIMILARITY"] = 'Similarity Score ' + df5.head(value_compound)["SIMILARITY"].round(2).astype(str)
df5.reset_index(inplace=True)
# df4 = df4.rename(columns={'Protein': 'symbol2'})
# print(df4)
# # Use df.query to get a subset of df1 based on ids in df2
# subset = df4.head(value_gene).query('symbol2 in @df2b.symbol2')
# # Use merge to join the two DataFrames on id
# result = pd.merge(subset, df2b, on='symbol2')
# print(result)
if value_compound <= df_len:
# Define the `text` column for labels and `href` column for links
# Reset the index
df12.reset_index(inplace=True)
# Replace hyphens with spaces in the 'text' column
df12['Compounds'] = df12['Compounds'].str.replace('-', ' ')
# Set the 'text' column back as the index
df12.set_index('Compounds', inplace=True)
df12['text'] = df12.index
df12['href'] = [f'https://pubmed.ncbi.nlm.nih.gov/?term={database_name}%5Bmh%5D+NOT+review%5Bpt%5D' \
'+AND+english%5Bla%5D+AND+hasabstract+AND+1990:2022%5Bdp%5D+AND+' + c for c in df12['text']]
df12['href2'] = [f'https://en.wikipedia.org/wiki/' + c for c in df12['text']]
df12['href3'] = [f'https://www.genome.jp/entry/{compound_id}' for compound_id in
get_compound_ids(df12['text'])]
assert isinstance(df12, object)
df12['database'] = database_name
# df11['name'] = [c for c in result['Approved name']]
# Create the treemap using `px.treemap`
fig = px.treemap(df12, path=[df12['text']], values=sizes,
custom_data=['href', 'database', 'href2', 'text', 'href3'],
hover_name=(df5.head(value_compound)['SIMILARITY']))
fig.update(layout_coloraxis_showscale=False)
fig.update_layout(autosize=True, paper_bgcolor="#CCFFFF", margin=dict(t=0, b=0, l=0, r=0))
fig.update_annotations(visible=False)
fig.update_traces(marker=dict(cornerradius=5), root_color="#CCFFFF", hovertemplate=None,
hoverlabel_bgcolor="lightblue", hoverlabel_bordercolor="#000000",
texttemplate="%{customdata[3]}
"
"PubMed"
"
Wikipedia"
"
KEGG Compound Page"
"")
fig.update_layout(uniformtext=dict(minsize=15), treemapcolorway=["LightYellow"])
# # display the treemap in Streamlit
# with treemap2:
# st.pyplot(fig2)
st.plotly_chart(fig, use_container_width=True)
st.caption("Compound designation and database provided by KEGG: https://www.kegg.jp/kegg/compound/")
csv = df1.head(value_compound).to_csv().encode('utf-8')
st.download_button(label=f"download top {value_compound} compounds (csv)", data=csv,
file_name=f'{database_name}_compounds.csv', mime='text/csv')
else:
st.warning(
f"This selection exceeds the number of similar proteins related to {query} within the {database_name} corpus, please choose a lower number")
# try:
# value_compound = min(df_len, 50)
# top_words = model.wv.most_similar_cosmul(query, topn=value_compound)
# words = df12.head(value_compound).index
# words = [word.replace(' ', '-') for word in words]
#
# sims = df5.head(value_compound)["SIMILARITY"].tolist()
# # print(sims)
# X_top = model.wv[words] # print(X_top)
# except Exception as e:
# print("Error:", e)
#
# # Remove the text "Similarity Score" from each element in the sims list
# sims_query_top = [float(sim.split()[-1]) for sim in sims]
# # print(sims_query_top)
#
# # Generate a 3D scatter plot of word embeddings using Plotly
# fig5 = px.scatter_3d(x=X_top[:, 0], y=X_top[:, 1], z=X_top[:, 2], text=words, color=sims_query_top,
# color_continuous_scale="RdYlGn", hover_name=words, hover_data={"color": sims_query_top})
#
# # Change background color to black
# fig5.update_layout(scene=dict(bgcolor='#CCFFFF'))
#
# # Change color of text to white
# fig5.update_layout(scene=dict(xaxis=dict(backgroundcolor='#CCFFFF', color='blue'),
# yaxis=dict(backgroundcolor='#CCFFFF', color='blue'),
# zaxis=dict(backgroundcolor='#CCFFFF', color='blue')))
#
# fig5.update_traces(hovertemplate='%{hovertext}
Similarity score: %{customdata[0]:.2f}')
# fig5.update_layout(
# title=dict(text=f"Word embedding map for {query} in {database_name} PubMed corpus", x=0.5, y=0.95,
# xanchor='center', yanchor='top', font=dict(color='black')),
# scene=dict(xaxis_title="Dimension 1", yaxis_title="Dimension 2", zaxis_title="Dimension 3"))
# fig5.update_coloraxes(colorbar_title="Similarity with query")
#
# # Represent query as a large red diamond
# fig5.add_trace(
# go.Scatter3d(x=[model.wv[query][0]], y=[model.wv[query][1]], z=[model.wv[query][2]], mode='markers',
# marker=dict(size=7, color='black', symbol='diamond'), name=query, hovertext=query,
# showlegend=False))
#
# # Add label for the query above the diamond
# fig5.add_trace(go.Scatter3d(x=[model.wv[query][0]], y=[model.wv[query][1]], z=[model.wv[query][2]], mode='text',
# text=[query], textposition='bottom center', textfont=dict(color='blue', size=10),
# hoverinfo='none', showlegend=False))
#
# # Add circles for the top 50 similar words
# fig5.add_trace(go.Scatter3d(x=X_top[:, 0], y=X_top[:, 1], z=X_top[:, 2], mode='markers',
# marker=dict(size=2, color=sims_query_top, colorscale='RdYlGn', symbol='circle'),
# hovertemplate='%{text}
Similarity score: %{customdata[0]:.2f}',
# text=words, customdata=sims, name=''))
#
# fig5.update(layout_coloraxis_showscale=True)
# fig5.update_layout(autosize=True, paper_bgcolor="#CCFFFF", margin=dict(t=0, b=0, l=0, r=0))
# fig5.update_annotations(visible=False)
#
# st.plotly_chart(fig5, use_container_width=True)
# st.markdown("---")
# import os
# from datasets import Dataset
# # Check if the comments directory exists
# if os.path.exists('comments'):
# # Load the dataset from disk
# dataset = Dataset.load_from_disk('comments')
# else:
# # Create a new dataset
# dataset = Dataset.from_dict({'id': [], 'text': []})
# def save_comment(comment):
# # Check if the dataset exists
# if os.path.exists('comments'):
# dataset = Dataset.load_from_disk('comments')
# else:
# dataset = Dataset.from_dict({'id': [], 'text': []})
# # Append the new comment to the dataset
# new_comment = {'id': len(dataset), 'text': comment}
# dataset = dataset.concatenate(Dataset.from_dict(new_comment))
# # Save the dataset to disk
# dataset.save_to_disk('comments')
# print('Comment saved to dataset.')
# st.title("Abstractalytics Web App")
# st.write("We appreciate your feedback!")
# user_comment = st.text_area("Please send us your anonymous remarks/suggestions about the Abstractalytics Web App: "
# "(app will pause while we save your comments)")
# if st.button("Submit"):
# if user_comment:
# save_comment(user_comment)
# st.success("Your comment has been saved. Thank you for your feedback!")
# else:
# st.warning("Please enter a comment before submitting.")
# # Load the comments dataset from disk
# if os.path.exists('comments'):
# dataset = Dataset.load_from_disk('comments')
# else:
# dataset = Dataset.from_dict({'id': [], 'text': []})
# # Access the text column of the dataset
# comments = dataset['text']
# # Define the password
# PASSWORD = 'ram100pass'
# # Prompt the user for the password
# password = st.text_input('Password:', type='password')
# # Display the comments if the password is correct
# if password == PASSWORD:
# st.title('Comments')
# for comment in comments:
# st.write(comment)
# else:
# st.warning('Incorrect password')
st.markdown("---")
except:
st.warning("")
st.subheader("Cancer-related videos")
if query:
idlist = []
search_keyword = {query}
html = urllib.request.urlopen("https://www.youtube.com/@NCIgov/search?query=cancer")
html2 = urllib.request.urlopen("https://www.youtube.com/@CancerCenter/search?query=cancer")
html3 = urllib.request.urlopen("https://www.youtube.com/@NorthwesternMedicine/search?query=cancer")
html4 = urllib.request.urlopen("https://www.youtube.com/@TEDEd/search?query=cancer")
html5 = urllib.request.urlopen("https://www.youtube.com/@CancerResearchUK/search?query=cancer")
video_ids = re.findall(r"watch\?v=(\S{11})", html.read().decode())
video_ids2 = re.findall(r"watch\?v=(\S{11})", html2.read().decode())
video_ids3 = re.findall(r"watch\?v=(\S{11})", html3.read().decode())
video_ids4 = re.findall(r"watch\?v=(\S{11})", html4.read().decode())
video_ids5 = re.findall(r"watch\?v=(\S{11})", html5.read().decode())
for i in video_ids2:
video_ids.append(i)
for i in video_ids3:
video_ids.append(i)
for i in video_ids4:
video_ids.append(i)
for i in video_ids5:
video_ids.append(i)
random.shuffle(video_ids)
c1, c2, c3 = st.columns(3)
with c1:
st.video("https://www.youtube.com/watch?v=" + video_ids[0])
with c2:
st.video("https://www.youtube.com/watch?v=" + video_ids[1])
with c3:
st.video("https://www.youtube.com/watch?v=" + video_ids[2])
st.markdown("---")
# Add a section header for useful resources
st.header("Learn More About Word2Vec the algorithm behind OncoDigger")
# Add links to videos and webpages
# Add links to videos and webpages
st.markdown("""
Here are some useful resources to help you learn more about Word2Vec:
1. [Word2Vec Tutorial - The Skip-Gram Model](http://mccormickml.com/2016/04/19/word2vec-tutorial-the-skip-gram-model/) - A blog post by Chris McCormick providing a detailed explanation of the skip-gram model used in Word2Vec.
2. [Word2Vec Tutorial Part 2 - Negative Sampling](http://mccormickml.com/2017/01/11/word2vec-tutorial-part-2-negative-sampling/) - A follow-up blog post by Chris McCormick discussing negative sampling in Word2Vec.
3. [Efficient Estimation of Word Representations in Vector Space](https://arxiv.org/pdf/1301.3781.pdf) - The original research paper by Mikolov et al. that introduced the Word2Vec algorithm.
4. [Word2Vec Tutorial: Vector Representation of Words](https://www.youtube.com/watch?v=64qSgA66P-8) - A YouTube video by Sentdex explaining the Word2Vec algorithm and its implementation in Python.
5. [Word2Vec: How to Implement Word2Vec in Python](https://www.youtube.com/watch?v=ISPId9Lhc1g&t=6s) - A YouTube video by Data Talks demonstrating how to implement Word2Vec in Python using the Gensim library.
6. [Cosine Similarity Calculator](https://www.omnicalculator.com/math/cosine-similarity) - A calculator for computing cosine similarity, a common metric used in measuring similarity between vectors.
""")
# else:
# st.error("The password you entered is incorrect.")