Is Brute Force the Last Resort of the Incompetent?

At a recent job interview, I was asked the question, "what are some ways you can think of to improve the brute force techniques used in drug discovery?" The question had an implicit bias, that brute force approaches to solving problems are less preferable to a well thought out alternative.

The same bias is built into the philosophy of this school. At MIT, a seldom spoken set of rules guides the actions of some of the most industrious nocturnal students. The Hacking Ethics is a set of 13 rules that encourage safety and respectful behavior of hackers, but the philosophy which crafted it also applies to real life. My favorite Ethic is rule 3:
"Brute force is the last resort of the incompetent."

This has a clear important application to hacking: don't break a window when you can open a door. There's often a smarter strategy than using force to accomplish your goal. This conventional wisdom applies to many aspects of life, but does it apply to pharmaceutical research? Is it necessarily true that brute force strategies should be the last resort in drug discovery?

I believe that in drug discovery, brute force techniques can be both cost and time effective, and often offer scientists a breadth of data they would not be able to obtain through other means. For example, in Module 2, we used small-molecule microarray slides to perform high-throughput screens of small molecules against a target protein. In this assay, molecules were tested for their binding affinities to our target. The SMM screen did not tell us the qualities of a small molecule that led to improved binding; it did not output where on the protein each small molecule was binding; nor did it predict on- or off-rate of the ligand. The SMM assay is a simple test of "binding" or "not binding", but that simplicity allows it to simultaneously test tens of thousands of molecules. This is a brute force method, and it's fast and cost effective at filtering through a large swath of candidates.

Virtual screening offers high-throughput screening methods at even larger orders of magnitude. During an internship at Relay Therapeutics, I built a system that allowed for docking simulations of 1 billion + small molecule candidates. These ultra-high-throughput virtual screening methods take days to run and can cost tens of thousands of dollars to secure time the many servers that run the software in parallel. This may seem like a large commitment, but it's actually fast and low-cost compared to many techniques used in the pharmaceutical industry. Additionally, for these virtual methods, the complexity of the output can be adjusted to gain more information from the simulations you are running, at the trade-off of taking more time or costing more money. Brute-force methods of simulating the activity of many molecules in a computer may not be a nuanced way of finding drug candidates, but it allows scientists to test a wider range of chemicals than they could ever get their hands on at one time, it gives them a set of data that would be difficult and expensive to obtain from laboratory techniques, and it acts as a powerful first-pass set of tests to accelerate the drug discovery process.

The most important rule in the Hacking Ethics reads, "Above all, exercise common sense." When you're doing drug discovery, ask: is there a brute force method that's easy, fast, and cost-efficient? If the answer's yes, then brute force should be your first resort, not your last.