My Take on Scientific Due Diligence

Scientific due diligence has been a major part of my work in biotech investing. Alongside the team and the business plan, the science is one of the three pillars we evaluate, and the one people ask me about most. How can you tell whether something is real science or snake oil?

My approach is to separate three things that are often blurred together: biological plausibility, the evidence itself, and the interpretation of that evidence.

Biological plausibility

It is relatively easy to build a plausible mechanism, so a good story proves very little on its own. There is some truth in every grift, so a grift usually involves stretching biological plausibility.

Evidence

Evidence is just what has actually been shown. In true frontier science, the goal is to identify clearly what you know and what you do not. You will have incomplete evidence, and that is fine. The useful exercise is to name the missing pieces, the experiments that would still have to be run for the whole idea to hold.

Interpretation

Most experiments carry important nuance, so it is easy to overstate what the results actually show. Interpretation has to stay anchored to the evidence. Problems begin when interpretation relies instead on biological plausibility or extends a finding to populations, outcomes, or settings in which it was never tested.

When a treatment fails to deliver, three common moves are often used to rescue the claim:

• Personalization. The treatment works, but only in people with the right biology.
• Time scale. The benefit is real, but it may take decades to become visible.
• Measurement. The treatment works, but our current tools are not sensitive enough to detect the effect in humans.

Each of these explanations is legitimate on its own, but they can be used as escape hatches to explain away every negative result. Together, they can make a claim impossible to disprove, producing an essentially unfalsifiable theory.

A truly scientific claim must be vulnerable to evidence that could sink it. Karl Popper argued that a theory compatible with every possible observation is unfalsifiable and therefore unscientific.¹ So pay close attention when people reach for these explanations to dismiss negative results.

That is ultimately how I distinguish frontier science from snake oil. Frontier science is explicit about what it does not know, identifies the evidence that would change its conclusions, and becomes more precise when results fall short. Most importantly, science allows reality to prove it wrong.

Footnotes

1. Karl Popper, The Logic of Scientific Discovery (1959). ↩