The second 'set' of your provided sample values contains 11 values. Is this a copy and paste error or intentional?

btw. you can provide split with a third argument to limit the number of values you want. So instead of using natatime() you could write:

while (my $line = <IN>) {
  ...
  my $remains = $line;
  while ($remains) {
    my (@set, $remains) = split /\t/, $remains, 11;
    # do something with 'sum(@set)'
  }
}
[download]

If there are less then 11 values left, then $remains will be set to undef and the inner loop terminates on the next iteration. The last set may contain less then 10 values - split does not complain, since the provided count value is just an upper limit.