python: association

Conda env

source /public/home/zzumgg03/huty/softwares/miniconda3/etc/profile.d/conda.sh
conda activate py37

association

#!/usr/bin/env python3
import pandas as pd
from scipy.stats import kendalltau,pearsonr,spearmanr

def kendall_pval(x,y):
    return kendalltau(x,y)[1]

def pearsonr_pval(x,y):
    return pearsonr(x,y)[1]

def spearmanr_pval(x,y):
    return spearmanr(x,y)[1]

datafile = "merge_out.txt"
df = pd.read_table(datafile, index_col=0)
df = pd.DataFrame(df.values.T, index=df.columns, columns=df.index)

#df = df.iloc[0:10,0:10]
co_r = df.corr(method = 'spearman')
co_p = df.corr(method = spearmanr_pval)

#
co_r.to_csv('./co_r.txt', sep='\t', na_rep='NA')
co_p.to_csv('./co_p.txt', sep='\t', na_rep='NA')

python实现R d_cast

#!/usr/bin/env python3
import pandas as pd
import os, sys

ms, infile, outfile = sys.argv

df = pd.read_csv(infile, index_col = 0, header = 0, sep = "\t")
with open(outfile, 'w') as o:
    for i in range(len(df.index)):
        for j in range(len(df.columns)):
            o.write("{}\t{}\t{}\n".format(df.index[i],df.columns[j],df.iloc[i, j]))

nohup python3 sc_melt.py co_p.txt co_p.melt &
nohup python3 sc_melt.py co_r.txt co_r.melt &

linux筛选

# 删除重复列
cat co_merge.melt | awk -F"\t" 'BEGIN{OFS="\t"}{$4=null;$5=null;print $0}' | sed 's/\t\t\t/\t/' > co_merge2.melt

# r > 0.8
cat co_merge2.melt | awk -F"\t" 'BEGIN{OFS="\t"}{if($3 >= 0.8) print $0}' > co_merge2_r0.8.melt

# bonf < 0.05/75325041 = 6.637899e-10
cat co_merge2_r0.8.melt | awk -F"\t" 'BEGIN{OFS="\t"}{if($4 < 6.637899e-10) print $0}' > co_merge2_r0.8_bonf.melt

# r > 0.9
cat co_merge2_r0.8_bonf.melt | awk -F"\t" 'BEGIN{OFS="\t"}{if($3 >= 0.99) print $0}' > 45万对相关

用 Python 对数据进行相关性分析
如何计算 Pandas 中列的成对相关的 p 值？