Ideas

It’s almost cliché to say that ideas are cheap. However, it’s an oversimplification—ideas live in a specific moment and within that context can have enormous or negligible impact. Ideas must also be implemented to have value.

I started keeping an idea notebook around 2015 to add some structure to the process of choosing new research directions. Almost everything I now work on once lived in this notebook, but the vast majority of ideas never make it out. I’ve decided to start collecting some of them here, sometimes with a short explanation of their history.

Screening for molecules that disrupt Gas6-PS interaction

We actually made an attempt at investigating this. The TAM receptors are a receptor tyrosine kinase family whose protein ligands bind to the lipid phosphatidylserine (PS). Current TAM inhibitors include decoy receptors, blocking antibodies, and kinase inhibitors, but these inhibit the receptors in a very different way than they are endogenously activated. A molecule that bound to Gas6 within its SBHD domain would block PS interaction and inhibit PS-induced activation. This would be a great tool compound for studying the TAM receptors, and perhaps also an interesting therapeutic approach.

Are haploinsufficient genes enriched in signaling factors?

This arose out of a discussion about how phosphorylation signaling is difficult to study due to nonlinear and highly dose-sensitive effects. A two-fold change in the level of many surface receptors, for example, can considerably change cell behavior. If this is generally the case, then it probably would also be true that these protein’s genes are more sensitive to copy number variation.

Can we make “ortho-antibodies”?

This also arose during a discussion, after a conference talk. CAR-T cells are in essence immune cells with artificial receptors directing their function. By introducing exogenous cells, we can induce an immune response otherwise inaccessible by protein or small molecule therapies. A key part of this is the introduced receptor, which targets those immune cells. The other set of proteins that are extremely variable within the body, besides the T cell receptor, are antibodies. So, what if one made antibodies that only interacted with an exogenously added set of immune cells? While an exciting idea, I couldn’t come up with cases where this approach would be uniquely valuable.

Can we make antibodies that get inside of cells?

Antibodies can be carried inside of cells during viral entry and can block viral replication and mount an intracellular immune response (via TRIM21). So, what if we made antibodies that recognized intracellular proteins? This is an idea that other groups have explored.