main.py

import pandas as pd
import networkx as nx
import numpy as np

# Read dataset (CSV)
# df_interact = pd.read_csv('data/processed_drug_interactions.csv')
single_drug_adr = pd.read_csv('data/Single_drug_adr.csv',index_col=0)
ddi = pd.read_csv('data/drug_drug.csv',index_col=0)
expert = pd.read_csv('data/expert_list.csv',index_col=0)
dpi = pd.read_csv('data/drug_gene_all.csv')
ppi = pd.read_csv('data/gene_gene.csv')

ncbi2name = pd.read_csv("data/ncbi2name.txt",sep='\t')
expert = pd.merge(expert, ncbi2name, left_on=['NCBI_ID'], right_on=['NCBI Gene ID(supplied by NCBI)'])
expert2 = expert[['NCBI_ID','Drug IDs','Drug IDs','Drug IDs','Approved symbol']]
expert2.columns = dpi.columns
dpi_L = pd.concat([dpi,expert2])
dpi_L = dpi_L.drop_duplicates(subset=['NCBI_ID','DrugBank ID_split'])

# MICROMEDEX_SEV_LEVEL

# Define list of selection options and sort alphabetically
drug_list = list(np.unique(single_drug_adr['source']))
drug_list.sort()

# Implement multiselect dropdown menu for option selection (returns a list)
selected_drugs = [drug_list[100],drug_list[200]]

# Create network graph when user selects >= 1 item

df_select = single_drug_adr.loc[single_drug_adr['source'].isin(selected_drugs)]
df_select = df_select.reset_index(drop=True)

ddi_select = ddi.loc[ddi['DRUG_1_CONCEPT_NAME'].isin(selected_drugs) | \
                            ddi['DRUG_2_CONCEPT_NAME'].isin(selected_drugs)]
ddi_select = ddi_select.reset_index(drop=True)

dpi_select = dpi.loc[dpi['drug_node_name'].isin(selected_drugs)]
dpi_select = dpi_select.reset_index(drop=True)
# dpi_select = dpi_select.groupby('Gene Name').filter(lambda x: len(x) > 1)

selected_genes = list(dpi_select['Gene Name'])
ppi_select = ppi.loc[ppi['source'].isin(selected_genes)| \
                            ppi['target'].isin(selected_genes)]
ppi_select = ppi_select.reset_index(drop=True)
#let first column be the common targets
ppi_select[['source','target']] = ppi_select[['target','source']].where( \
    ppi_select['target'].isin(selected_genes), ppi_select[['source','target']].values)

# Create networkx graph object from pandas dataframe
G = nx.from_pandas_edgelist(df_select, 'source', 'target', 'rel')

dpi_select.columns = ['source','target']
L = pd.concat([dpi_select,ppi_select])
G = nx.from_pandas_edgelist(L, 'source', 'target')
# all shortest path
print([p for p in nx.all_shortest_paths(G, source=drug_list[100], target=drug_list[200])])
ps = [p for p in nx.all_shortest_paths(G, source=drug_list[100], target=drug_list[200])]
ls = ps[0]
rest = ls[1:-1]
# rest = ps.sort(key=len)
# ps_DF = pd.DataFrame(ps).iloc[:, 1:-1]
print('--')