This will let you specify a "Base" column as the first argument on the command line, followed by any number of other columns to count and sum (the first column is numbered zero, as in your example):

abaugher@cail> cat 1062685.pl
#!/usr/bin/env perl
use strict; use warnings;

my $key = shift;
my @cols = @ARGV;
my %h;

die "Usage: 1062685.pl id_column field1 [field2]...\n" unless @cols;
# $ARGV[0] is the Base column
# $ARGV[1..x] is the list of columns to add up

while(<DATA>){
  chomp;
  my @f = split;
  for (@cols){
    $h{$f[$key]}{$_}{t} += $f[$_];
    $h{$f[$key]}{$_}{n}++;
  }
}

for my $k (sort keys %h){
  print $k;
  print "\t$h{$k}{$_}{n}\t$h{$k}{$_}{t}" for @cols;
  print "\n";
}

__DATA__
U1     ID1     100     280
U1     ID1     137     250
U2     ID2     150     375
U1     ID2     100     100
U3     ID1     100     600
U9     ID3     137     200

abaugher@cail> perl 1062685.pl 1 2 3               
ID1     3       337     3       1130
ID2     2       250     2       475
ID3     1       137     1       200
abaugher@cail> perl 1062685.pl 0 2 3                   
U1      3       337     3       630
U2      1       150     1       375
U3      1       100     1       600
U9      1       137     1       200
[download]