代码思路

python

#!/usr/bin/env python3


import argparse


if __name__ == '__main__':
    parser = argparse.ArgumentParser(description='test....')
    parser.add_argument('-p', '--pdb', required=True, type=str,
                        help='the pdb id of protein')
    parser.add_argument('-c', '--chain', default='A', type=str,
                        help='the chain of protein, Default: %(default)s')
    parser.add_argument('--num', required=True, type=int,
                        help='the index of atom')
    args = parser.parse_args()


    print(f"pdb id is {args.pdb}")
    print(f"the chain of protein is {args.chain}")
    print(f"the index of atom is {args.num}")

GPU

### 查看GPU使用情况
nvidia-smi
### 指定某块GPU（默认0盘）
export CUDA_VISIBLE_DEVICES=0
export CUDA_VISIBLE_DEVICES=1
export CUDA_VISIBLE_DEVICES=2

shell文本操作

shuf

shuf file.txt 

grep

awk

数据清洗

# 获取某一列非空的数据
awk -F ',' '{if($2!="")print $0}' 1.txt


# 取两位小数（会把字符串转为0，故建议awk处理浮点数列，结合paste整合字符串列）
# 下面这个例子是把第三列后的全部取两位小数
awk '{for(i = 3; i <= NF; i++) {printf("%.2f\t", $i)} {printf("\n")}}' analysis.txt


# OFS 指定输出的分隔符
sort -t , -r -n -k 2 n2-mf.csv|awk -v OFS="," -F ',' '{print $1,$2}'|head >> n2-count.txt


# 使用 awk 根据特定列去重 （gpt)
awk '!seen[$2]++' data.txt

数据统计

# 获取第3列数据中大于1的数据行
awk '$3<-1 {print}' 1.txt


# 用于MMGBSA 计算后统计结合能大于-1kcal/mol的残基
awk -F ' ' '$28<-1 {print}' FINAL_DECOMP_MMPBSA.dat
for i in {1,2,28};do awk -F ' ' '$28<-1 {print $'${i}'}' 400-500ns_c4_atps-site/FINAL_DECOMP_MMPBSA.dat;done


# 功能类似于grep，但更灵活
awk -F ';' '$10=="寒" {print $1,$10}' ~/zzy/origin/230914_tcm_label.csv >>hl.txt


# 按列做乘法并輸出 
awk -F ',' '{print $1*10,$2*10,$3}' ref_pro_mol_rmsd.csv > ref_pro_mol_rmsd_A.csv 


awk '{print $1*0.2}' num|awk -F '.' '{print $1}' >0.2num


# 输出每行数据的最大值
awk '{for(i = 2; i <= NF; i++) {if($i < $(i - 1)) {$i = $(i - 1)}} {print $NF}}' test.txt




# 变量计算
top1_rate=$(awk 'BEGIN{print "'$top1'"/"'$sum'"}')
sum_higher_rate=$(awk 'BEGIN{print "'${sum_higher_num}'"/"'${all_num}'"}')
score_frag_sum=$(awk 'BEGIN{print "'${score_frag_i_1}'"+"'${score_frag_i_2}'"}')




all_num=`expr ${all_num} + 1`
sum_higher_num=`expr ${sum_higher_num} + 1`

paste

paste -d ',' file temp > input_list.csv

sed

### Print odd lines (~2 means skip 2 lines)
sed -n '1~2p'


### 
sed -i 's/oo/kk/g' testfile
sed -i 's/Performance:  //g' 


###   
sed -i "2c $i" $i.mol2

text manipulation

grep -A 1 min dock-1_c1.sdf |sed -n '/-/p'|sed -n '1~2p'
 
### get lines that exist in a but are missing in b.
cat a b b | sort | uniq -u


### Convert plain text files from DOS or Mac format to Unix
dos2unix *


### practical
for a in {1..51};do b=`sed -n "${a}p" SW`; sed -i "1i ${a}.sdf" $b;done 
for a in {1..10};do sed -i "{$a/$/$a/}" test.smi;done 


### vi replace ：
:1,$s/\[/\\[/g


:12,22s/abc/123/g 


# for 循环中用文本文件，空格会识别为分隔符，可以先在vi里用特殊符号替换空格，再用sed将$i 中的特殊符号替换成空格。由此，可以正确grep
for i in `cat up1k_target_hl.txt`;do b=`echo $i|sed 's/_/ /g'`;grep -w "$b" /home/yqyang/DDB-AI/script/Index_MPNN_CNN.txt >>id.txt;done