I have two sam files produced by alignment to two different versions of the reference genome. Is there a way to get the intersection of the sam files so I get a sam with the sequences that have been mapped to both? (So, I would ideally want to discard any sequences that are unmapped in an alignment to at least one of the genomes.)