Vivek Raj Senthivel

A central mistake in biology was to name genes. This over-simplification made reconciling what is happening on the molecular level a mess - it's not rare to find reports of opposite mechanisms in different contexts, claimed involvement in dozens if not hundreds of different processes, sometimes inhibiting and sometimes amplifying and most of the time being oblivious to the potential for sequence-level variation. Nobody would be surprised about this diversity of findings if we instead recognized genes as (sometimes quite lengthy and complex) pieces of sequence that carry state and interact with and are interpreted by their environment - often producing dozens of gene products that are in turn themselves context-dependent and modulated. Naturally, such a highly amorphous composition of objects has many diverse effects, and masking this complexity behind a single name more often than not ends up being a harmful abstraction. The primary role of gene names then is to give us the false appearance of comfort in the face of enormous biological complexity. One under-appreciated potential of the emergence of AI tools in biology is to undo this mistake, and - instead of assuming it away - extend our ability to lean further into this complexity.