If you're only interested in processing the ATOM records, then this may be good enough:

#! perl -slw
use strict;
use List::Util qw[ sum ];

my %stats;

while( <DATA> ) {
    next unless m[^ATOM];
    my @fields = unpack 'a6 a5 a3 a3 a4 x2 a3 x4 a8 a8 a7 a7 a6 a12', 
+$_;
#    print join'|', @fields;
    push @{ $stats{ $fields[ 4 ] } }, $fields[ 10 ];
}

printf "%3s : %.3f%%\n", $_,
    sum( @{ $stats{ $_ } } ) / @{ $stats{ $_ } }
    for sort keys %stats;

=output
C:\test>junk2
GLN : 35.962%
LEU : 25.250%
LYS : 64.087%
=cut

__DATA__
1234567890123456789012345678901234567890123456789012345678901234567890
+1234567890
ATOM      1  N   LYS A   2     -10.231   0.078  18.128  1.00 28.52    
+       N
ATOM      2  CA  LYS A   2      -9.472  -0.327  19.297  1.00 28.52    
+       C
ATOM      3  C   LYS A   2     -10.308  -0.875  20.459  1.00 28.52    
+       C
ATOM      4  O   LYS A   2     -10.180  -2.054  20.801  1.00 28.52    
+       O
ATOM      5  CB  LYS A   2      -8.592   0.833  19.765  1.00 92.54    
+       C
ATOM      6  CG  LYS A   2      -7.346   0.373  20.498  1.00 92.54    
+       C
ATOM      7  CD  LYS A   2      -6.207   1.375  20.382  1.00 92.54    
+       C
ATOM      8  CE  LYS A   2      -4.929   0.810  21.002  1.00 92.54    
+       C
ATOM      9  NZ  LYS A   2      -3.756   1.724  20.882  1.00 92.54    
+       N
ATOM     10  N   GLN A   3     -11.162  -0.038  21.053  1.00 28.24    
+       N
ATOM     11  CA  GLN A   3     -11.998  -0.465  22.181  1.00 28.24    
+       C
ATOM     12  C   GLN A   3     -12.749  -1.750  21.849  1.00 28.24    
+       C
ATOM     13  O   GLN A   3     -12.882  -2.639  22.694  1.00 28.24    
+       O
ATOM     14  CB  GLN A   3     -13.019   0.618  22.556  1.00 42.14    
+       C
ATOM     15  CG  GLN A   3     -12.450   2.018  22.718  1.00 42.14    
+       C
ATOM     16  CD  GLN A   3     -12.541   2.849  21.437  1.00 42.14    
+       C
ATOM     17  OE1 GLN A   3     -12.042   2.453  20.377  1.00 42.14    
+       O
ATOM     18  NE2 GLN A   3     -13.180   4.012  21.536  1.00 42.14    
+       N
ATOM      1  N   LEU A  62      32.071  22.080  63.175  1.00 25.25    
+       N
ATOM      2  CA  LEU A  62      32.171  21.102  62.056  1.00 25.25    
+       C
ATOM      3  C   LEU A  62      31.367  19.843  62.260  1.00 25.25    
+       C
ATOM      4  O   LEU A  62      31.676  18.842  61.650  1.00 25.25    
+       O
ATOM      5  CB  LEU A  62      33.640  20.698  61.795  1.00 25.25    
+       C
ATOM      6  CG  LEU A  62      34.589  21.606  60.965  1.00 25.25    
+       C
ATOM      7  CD1 LEU A  62      35.877  20.829  60.616  1.00 25.25    
+       C
ATOM      8  CD2 LEU A  62      33.903  22.137  59.675  1.00 25.25    
+       C
ATOM      9  N   LEU A  63      30.358  19.879  63.124  1.00 25.25    
+       N
ATOM     10  CA  LEU A  63      29.497  18.715  63.396  1.00 25.25    
+       C
ATOM     11  C   LEU A  63      28.036  19.160  63.458  1.00 25.25    
+       C
ATOM     12  O   LEU A  63      27.574  19.700  64.489  1.00 25.25    
+       O
ATOM     13  CB  LEU A  63      29.869  17.995  64.717  1.00 25.25    
+       C
ATOM     14  CG  LEU A  63      29.125  16.690  65.118  1.00 25.25    
+       C
ATOM     15  CD1 LEU A  63      29.725  15.515  64.383  1.00 25.25    
+       C
ATOM     16  CD2 LEU A  63      29.208  16.398  66.606  1.00 25.25    
+       C
[download]

Thanks for spending time to write the code, helped a lot. But is there a way for me to apply this to a big chunk of data ? I mean, the aminoacid groups go as long as 1,2,...............264 and there can be 2600something ATOM lines depending on the PDB file.

Sure. If you supply this version:

#! perl -slw
use strict;
use List::Util qw[ sum ];

BEGIN{ @ARGV = map glob( $_ ), @ARGV }

my %stats;

while( <> ) {
    next unless m[^ATOM];
    my @fields = unpack 'a6 a5 a3 a3 a4 x2 a3 x4 a8 a8 a7 a7 a6 a12', 
+$_;
#    print join'|', @fields;
    push @{ $stats{ $fields[ 4 ] } }, $fields[ 10 ];
}

printf "%3s : %.3f%%\n", $_,
    sum( @{ $stats{ $_ } } ) / @{ $stats{ $_ } }
    for sort keys %stats;
[download]

with a wildcard path/filename on the command line,

theScript.pl /path/to/*.pdb
[download]

it will perform the stats across all the matching files.

Alternatively, if you want the stats on a file-by-file basis, then supply the filenames one at a time:

## For windows; something similar is possible on *nix
for %i in (\path\to\*.pdb) do @theScript.pl \path\to\%i
[download]

---

Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.

"Science is about questioning the status quo. Questioning authority".

In the absence of evidence, opinion is indistinguishable from prejudice.

"I'd rather go naked than blow up my ass"

I got stuck in the regex

I would suggest starting with CPAN rather than writing your own parser. Bio::Structure::IO, Chemistry::File::PDB, Chemistry::Mol all look relevant, and there are likely several more. If you use an existing module to do the low-level work, you can focus on the real reason you are writing the code (i.e., it isn't because you want to parse a PDB file, it is because you want to do something with the contents of the file).

If you insist on rolling your own parser, look at unpack. It is great for fixed-width data, and will likely be much faster and less buggy than any regex you assemble.

HTH