I think at some point in a computer scientist’s studies the computers themselves become less of a thing, a tool, and instead become more of an extension of the self like an arm or a hand. I’ve always felt the true power of computers comes not in completely automated processes, but in combining the analytical strength of computers with the very powerful pattern recognition abilities of the human mind. To some extent my research is trending in that direction with the idea digital forensics needs to move away from a synchronous process to an asynchronous one. You can see a previous post summarizing my AAFS presentation on Black Friar.
As a slight background, I spent a couple semesters moonlighting as a research assistant with the bioinformatics group (Computer Science + Biochem). My brief stint of research was on Intron-Exon and Exon-Intron splice site sensors not protein folding, but I sat through several lectures, conversations, and a few conference presentations on it so I know how complex the topic is. This article caught my eye and essentially is doing in the area of protein folding that I suggest should be done in digital forensic analysis.
Check it out, it is a good read. Essentially, the computational algorithms are good at handling the short chains of amino acids, but loose precision as the structures grow. Humans are bad at handling short chains, but good at optimizing the folds in larger structures. Researchers tapped into that by making the protein folding screen saver a game.