基于复杂网络的药物疾病关系预测方法python实现源码.zip

# coding: utf-8 import sys import numpy as np import pandas as pd from sklearn import preprocessing import networkx as nx import xlrd import random sys.path.append("sideMain2.py") import sideMain2 df1 = pd.read_csv(r"E:\bigdata\data\du.csv") def gene_select(gene): row = int(df1[df1['ge'] == gene].index.values) # 该基因所在行 if row < 100: # 这个基因前或后100个基因中随机挑 g1 = df1.iloc[0:row + 100] elif row > 17605: g1 = df1.iloc[row - 100:17705] else: g1 = df1.iloc[row - 100:row + 100] g = int(g1.sample(n=1)['ge']) # 随机选取一个基因数据，并转为int类型 return g # 数据导入，预处理 edgelist = [] def open_excel(file): try: sheet_data = xlrd.open_workbook(file) return sheet_data except Exception as e: print(str(e)) def praise_excel(file): sheet_data = open_excel(file) sheet = sheet_data.sheets()[0] #读取sheet my_queues0 = sheet.col_values(0) #读取1列 my_queues1 = sheet.col_values(1) #读取2列 for i, j in zip(my_queues0,my_queues1): x = (int(i),int(j)) edgelist.append(x) praise_excel(r"E:\bigdata\data\Human Interactome_201706.xlsx") G=nx.Graph() #创建了一个没有节点和边的空图 #列表 G.add_edges_from(edgelist) #对于无向图，边3-2与边2-3被认为是一条边 model = sideMain2.xlsx_file.parse('Gene Vector', index_col='Gene') def zscore(): df = sideMain2.xlsx_file.parse('00Drug-Gene_Network') grouped = df.groupby('Drug') drug = [] target = [] for group in grouped.groups: drug.append(group) target.append(list(grouped.get_group(group)['Gene'])) diseaseGene = sideMain2.research_diseaseGene('breast neoplasms') disease_gene = [] for dG in diseaseGene: if np.int(dG) in model.iloc[:, 0]: disease_gene.append(np.int(dG)) closest = [] flg = 0 for item in target: # 2833 print(flg) flg += 1 closest1 = [] for i in range(1000): num = len(item) distance = 0.0 for ge in item: # 30药物 min_dis = 10 if ge in model.iloc[:, 0]: g1 = gene_select(ge) else: num -= 1 if num == 0: break else: continue for gene in disease_gene: g2 = gene_select(gene) # 提取本地基因与基因之间的边数 try: t = nx.shortest_path_length(G, g1, int(g2)) except nx.NetworkXNoPath: continue min_dis = min(min_dis, t) if min_dis == 10: num -= 1 else: distance = distance + min_dis if num == 0: closest1.append(np.NAN) else: closest1.append(distance / num) if num < 0: print("num<0") closest.append(closest1) df = pd.DataFrame(closest) df.to_csv('df_asthma.csv') if __name__ == "__main__": zscore() df = pd.read_csv('df_asthma.csv', index_col=0) result = pd.DataFrame({'mean': pd.Series(df.mean(axis=1)), 'std': pd.Series(df.std(axis=1))}) # result.to_csv('result_mean_std_asthma.csv') df1 = pd.read_csv('resultSide_breast neoplasms.csv') result = pd.concat([df1, result], axis=1) result.to_csv('6144-8515resultSide_breast neoplasms.csv') #df2 = pd.read_csv('resultSide_zscore_asthma.csv') #df2['zscore'] = (df2['distance'] - df2['mean']) / df2['std'] #df2 = df2.sort_values(['zscore']) # 升序 #df2.to_csv('resultSide_zscore_asthma.csv')