bash script

Question

[priyanka]

Hello everyone, I have a file like this: RSID1 RSID2

chr1_169894240_G_T_b38  chr1_169894240_G_T_b38
chr1_169894240_G_T_b38  chr1_169891332_G_A_b38
chr1_169891332_G_A_b38  chr1_169891332_G_A_b38
chr1_169661963_G_A_b38  chr1_169661963_G_A_b38
chr1_169661963_G_A_b38  chr1_169697456_A_T_b38
chr1_169697456_A_T_b38  chr1_169697456_A_T_b38
chr1_27636786_T_C_b38   chr1_27636786_T_C_b38
chr1_196651787_C_T_b38  chr1_196651787_C_T_b38
chr6_143501715_T_C_b38  chr6_143501715_T_C_b38

I want to extract info just like: chr1_169894240 chr1_169894240. I don't want to have other info. I just want chr_pos I am confuse how to extract this info because the length is varying. In one case its 9 length and in other its 10. So if i use cut command for some its showing write value like chr_pos but for some its showing chr_pos_ Can anyone please help me out with this.

Answer 1

[alex.zaccaron]

You can use cut or awk with "_" as field separator character, e.g., cut -f 1 yourfile.txt | awk -v FS="_" {print $1"_"$2}. If you have a 2-column tsv file, you can try:

paste <(cut -f 1 yourfile.txt  | awk -v FS="_" '{print $1"_"$2}') <(cut -f 2 yourfile.txt  | awk -v FS="_" '{print $1"_"$2}')

[priyanka]

Thank you so much. It worked. Can you also share the link where I can learn in detail about awk command. I know just the basic of it

Answer 2

[cpad0112]

$ sed -r 's/_\w_\w_\w{3}//g' test.txt

$ awk -v OFS="\t" -F '[_\t]' '{print $1"_"$2,$6"_"$7}' test.txt

$ parallel --colsep "_|\t" echo {1}_{2} {6}_{7} :::: test.txt  | sed 's/\s/\t/'

chr1_169894240  chr1_169894240
chr1_169894240  chr1_169891332
chr1_169891332  chr1_169891332
chr1_169661963  chr1_169661963
chr1_169661963  chr1_169697456
chr1_169697456  chr1_169697456
chr1_27636786   chr1_27636786
chr1_196651787  chr1_196651787
chr6_143501715  chr6_143501715

[priyanka]

Thank you. I am able to do it by this command also

Answer 3

[Kevin Blighe]

For the win, can even do a fancy regex with sed

cat data.tsv 
chr1_169894240_G_T_b38  chr1_169894240_G_T_b38
chr1_169894240_G_T_b38  chr1_169891332_G_A_b38
chr1_169891332_G_A_b38  chr1_169891332_G_A_b38
chr1_169661963_G_A_b38  chr1_169661963_G_A_b38
chr1_169661963_G_A_b38  chr1_169697456_A_T_b38
chr1_169697456_A_T_b38  chr1_169697456_A_T_b38
chr1_27636786_T_C_b38   chr1_27636786_T_C_b38
chr1_196651787_C_T_b38  chr1_196651787_C_T_b38
chr6_143501715_T_C_b38  chr6_143501715_T_C_b38

sed 's/_[ATGC]_[ATGC]_[a-z][0-9]*//g' data.tsv 
chr1_169894240  chr1_169894240
chr1_169894240  chr1_169891332
chr1_169891332  chr1_169891332
chr1_169661963  chr1_169661963
chr1_169661963  chr1_169697456
chr1_169697456  chr1_169697456
chr1_27636786   chr1_27636786
chr1_196651787  chr1_196651787
chr6_143501715  chr6_143501715

Kevin

[priyanka]

Thank you so much. this is quite easy to understand