Try to use the third argument to split like this

my( $chr, $start, $end, $description ) = split( /\s+/ , $annotation_li
+ne, 4 );
[download]

This way, $description will contain all of the remaining line and not only up to the next whitespace.