Filtering input

Perl's text processing features are ideally suited for filtering operations that extract the desired content from a file.

1. Skip empty lines and lines containing only white space:
   

   perl -lne 'print if (/\S/)' input > output
   
   

2. Filter on absolute fold-changes and p-values, located in this case in columns 3 and 5 (indexed 2 and 4), respectively:
   

   perl -lane 'print if (abs($F[2]) >= 2 and $F[4] <= 0.05)' input > output
   
   

3. Extract lines that contain Ensembl gene identifiers:
   

   perl -lne 'print if (/ENSG\d+/)' input > output
   
   

4. Extract lines that contain membrane-related terms, case-insensitive:
   

   perl -lne 'print if (/membrane/i)' input > output
   
   

5. Sub-sampling -- extract approximately 1 percent of random lines from an input file:
   

   perl -lne '$i = rand; print if ($i <= 0.01)' input > output