Keep shared entries among many files - bash

Question

I have hundreds of files, each with different number of entries (>xxxx) and want to keep only shared entries among all files, separately. I'm not sure what is the best method to do this, maybe perl! I used sort, uniq of bash, but I didn't get the correct answer. The format of IDs start with > and follows 4 characters among all files.
1.fa
>abcd
CTGAATGCC
2.fa
>abcd
AAATGCGCG
>efgh
CGTAC
3.fa
>abcd
ATGCAATA
>efgh
TAACGTAA
>ijkl
TGCAA
Final results, of this example would be:
1.fa
>abcd
CTGAATGCC
2.fa
>abcd
AAATGCGCG
3.fa
>abcd
ATGCAATA

Answer 1

This Perl program will do as you ask. It uses Perl's built-in edit in place functionality and renames the original files to 1.fa.bak etc. It shouldn't have a problem with blank lines in your data as long as the sequence is always on one line immediately following the ID
use strict;
use warnings 'all';
my #files = glob '*.fa';
printf "Processing %d file%s\n", scalar #files, #files == 1 ? "" : "s";
exit if #files < 2;
my %ids;
{
local #ARGV = #files;
while ( <> ) {
++$ids{$1} if /^>(\S+)/;
}
}
# remove keys that aren't in all files
delete #ids{ grep { $ids{$_} < #files } keys %ids };
my $n = keys %ids;
printf "%d ID%s common to all files\n", $n, $n == 1 ? '' : "s";
exit unless $n;
{
local #ARGV = #files;
local $^I = '.bak';
while ( <> ) {
next unless /^>(\S+)/ and $ids{$1};
print;
print scalar <>;
}
}

Answer 2

Here is Perl solution, that may help you:
use feature qw(say);
use strict;
use warnings;
my $file_dir = 'files';
chdir $file_dir;
my #files = <*.fa>;
my $num_files = scalar #files;
my %ids;
for my $file (#files) {
open ( my $fh, '<', $file) or die "Could not open file '$file': $!";
while (my $id = <$fh>) {
chomp $id;
chomp (my $sequence = <$fh>);
$ids{$id}++;
}
close $fh;
}
for my $file (#files) {
open ( my $fh, '<', $file) or die "Could not open file '$file': $!";
my $new_name = $file . '.new';
open ( my $fh_write, '>', $new_name ) or die "Could not open file '$new_name': $!";
while (my $id = <$fh>) {
chomp $id;
chomp (my $sequence = <$fh>);
if ( $ids{$id} == $num_files ) {
say $fh_write $id;
say $fh_write $sequence;
}
}
close $fh_write;
close $fh;
}
It assumes that all the .fa files are located in the directory named $file_dir, and it writes the new sequences to new files in the same directory. The new file names get the .new extension.