That approach only works because yeast has been subjected to intense investigation by low-throughput techniques, providing a huge knowledge base that constrains and guides the automated investigation. (It also helps that yeast doesn't do alternative splicing.) So it's not so much "replacing" as "building upon".