如何用sed命令替换特定字符的字符串

Question

I want to use the sed command to delete some specific strings.

This is the file(tRNA.fa):

>tRNA-Ala-AGC-1-1 (chrII.trna5-AlaAGC) chrII:4565386-4565457 (+) Ala (AGC) 72 bp Sc: 72.4
GGGGGTATAGCTCAGTGGTAGAGCGCTCCCTTAGCATGGGAGAGGgCTGGGGTTCAATTC
CCCATACCTCCA
>tRNA-Ala-AGC-1-10 (chrX.trna261-AlaAGC) chrX:7378738-7378809 (-) Ala (AGC) 72 bp Sc: 72.4
GGGGGTATAGCTCAGTGGTAGAGCGCTCCCTTAGCATGGGAGAGGgCTGGGGTTCAATTC
CCCATACCTCCA
>tRNA-Ala-AGC-1-11 (chrX.trna260-AlaAGC) chrX:7507245-7507316 (-) Ala (AGC) 72 bp Sc: 72.4
GGGGGTATAGCTCAGTGGTAGAGCGCTCCCTTAGCATGGGAGAGGgCTGGGGTTCAATTC
CCCATACCTCCA

I just want to keep “>tRNA-XXX-XXX-XX”and the next line.

So,I tried to replace unnecessary information with sed command:

sed -i 's/\(.*\).*[0-9]$//g' tRNA.fa

However,I deleted all the line starting with '>'.

The result I hope to get is:

>tRNA-Ala-AGC-1-1 
GGGGGTATAGCTCAGTGGTAGAGCGCTCCCTTAGCATGGGAGAGGgCTGGGGTTCAATTC
CCCATACCTCCA
>tRNA-Ala-AGC-1-10 
GGGGGTATAGCTCAGTGGTAGAGCGCTCCCTTAGCATGGGAGAGGgCTGGGGTTCAATTC
CCCATACCTCCA
>tRNA-Ala-AGC-1-11 
GGGGGTATAGCTCAGTGGTAGAGCGCTCCCTTAGCATGGGAGAGGgCTGGGGTTCAATTC
CCCATACCTCCA

If you know how to replace it, please tell me,thank you.

Answer 1

If you want to match a ( don't escape it.

sed -i 's/(.*).*[0-9]$//g' tRNA.fa

But really the following is just enough to remove everything after ( :

sed -i 's/(.*//' tRNA.fa

Note that you may also want s/ (.*// to remove that space before ( too.

The \\(...\\) is used to group expression into a subgroup. It is most often used for back references, but may also be used for example as \\(abc\\)* - will search zero or more occurrences of the string "abc" .

Here is a great sed introduction.