T检验

from scipy import stats

stats.ttest_1samp(data,1)

stats.ttest_ind(data1,data2)

stats.levene(data1,data2)

stats.ttest_ind(data1,data2,equal_var=False)

stats.ttest_rel(data1,data2)

#!/usr/bin/python
# -*- coding: utf-8 -*-


import pandas as pd
from scipy import stats
import re
import statsmodels.stats.multitest
import click
import numpy as np

@click.group()
def stat_tools():
  ‘‘‘
 This tool is for deal with GSE data,Protein data or Metabolism  data ‘‘‘
def get_pvalue(df,control_list,treat_list):
  p_result = stats.ttest_ind(df[control_list],df[treat_list],equal_var=False)
    pvalue = p_result.pvalue
    return pvalue

@click.command()
@click.option("-i","--input", help="a gene expression matrix file.",type=click.Path(exists=True))
@click.option("-o","--output", help="a gene expression matrix file with pvalue, qvalue and Foldchange.")
@click.option("-c","--control", help="control group name, sample name is like Normal_1 or Normal-1.")
@click.option("-t","--treat", help="Treat group name, sample name is like Tumor_1 or Tumor-1.")
@click.option("-f","--foldchange", default=1.5,help="foldchange to filter DEG result, default 1.5.")
@click.option("-p","--pvalue", default=0.05,help="pvalue to filter DEG result,default 0.05.")
@click.option("--filtered/--nofiltered", default=False,help="if create a gene expression matrix file filter by threshold.")
def Ttest(input,output,control,treat,foldchange,pvalue,filtered):
  ‘‘‘
 use T-test to get DEG from GSE data,Protein data or Metabolism  data ‘‘‘  df = pd.read_csv(input,sep="\t",header=0)

    control_list = [n for n in df.columns if re.match(control,n)]
    treat_list = [n for n in df.columns if re.match(treat,n)]

    df["pvalue"] = df.apply(get_pvalue,axis=1,args = (control_list,treat_list))
    df["log2FoldChange"] = df[treat_list].mean(axis=1) - df[control_list].mean(axis=1)
    df["qvalue"]=statsmodels.stats.multitest.multipletests(df["pvalue"],method=‘fdr_bh‘)[1]

    df.to_csv(output,sep="\t",index=False,header=True)

    if filtered:
  filter_name="DEGs_"+output.replace(".xls","")+".FC"+str(foldchange)+"_pvalue"+str(pvalue)+".xls"
  df_filter = df.loc[(df["pvalue"] <= pvalue ) & (np.abs(df["log2FoldChange"]) >=np.log2(foldchange))]
        df_filter.to_csv(filter_name,sep="\t",index=False,header=True)

stat_tools.add_command(Ttest)

if __name__ == ‘__main__‘:
  stat_tools()