#275 | Thomas Stoeger; Identifying gaps in human knowledge, aging, peer review

About this episode

Thomas Stoeger, a postdoctoral researcher at Northwestern University, joins Vance to discuss his research into the "gaps" in scientific literature — genes, proteins, and mechanisms that appear promising in large unbiased datasets but have never actually been studied in detail. Stoeger explains that roughly half of the most-promising Alzheimer's-related genes flagged by NIH have never been investigated, and argues this is a social/career phenomenon rather than a purely scientific one: junior researchers who study under-explored genes tend to advance more slowly, likely because they lack a "tribe" of peers, reference-letter writers, and conference community around a novel topic. He frames the dynamic as a "tournament economy" where being second in a crowded, prestigious field is worse than being best in an obscure one. The conversation broadens into a discussion of tribalism in science and public discourse generally — Vance draws a direct parallel to his own Monsanto experience, where being anti-GMO or pro-GMO becomes a group identity that people defend as "righteous" regardless of the underlying evidence, and the two discuss how threatened identity produces group-protective, emotionally charged reactions rather than reasoned rebuttal. They also touch on the Pareto distribution in gene research, the value of widening scientific inquiry versus the legitimate case for speed/first-mover advantage, and nutrition science as an area (following a recent guest, Laszlo Bassi) that may be similarly oversimplified around macronutrients while ignoring a vast number of unstudied food molecules.

“I think that many people maybe favor the first one over the second one and I think that is a problem for policy makers... basically just kind of peer review, dis review and basically other scientists that then discourage the NIH and other policy makers from exploring the unknown further because there's like more risk, there's more uncertainty.”

“There's like some arguments that can be said that if there's too many publications it'll actually slow down signs basically making it more difficult for new ideas to surface... sociologists of the University of Chicago have basically shown that in research fields that grow bigger and bigger... the turnover of ideas actually becomes slower.”

“Roughly half of the elements that seem to be most promising in this large amount of data sets have actually never been investigated by any scientist, which is a surprisingly large number... only half of them actually have been studied in greater detail by any scientists studying Alzheimer's disease.”

Key moments

• ~0-8%: Framing of Stoeger's core research question — using machine learning to spot genes/mechanisms that "should" be studied based on data patterns but have been ignored by the scientific literature.
• ~8-15%: The Alzheimer's example — roughly half of NIH-flagged promising genes have never been investigated in detail.
• ~15-27%: Why junior researchers who study under-explored genes progress more slowly in their careers — the "tournament economy" and tribal/community dynamics of science.
• ~40-46%: Pareto distribution in gene research — whether 20% of genes really account for 80%+ of biological importance, or whether that's an artifact of where scientists chose to look.
• ~85-92%: Direct parallel between anti-GMO/pro-GMO tribalism at Monsanto and the way threatened group identity produces "righteous" emotional defense rather than reasoned argument, plus the bootcamp/shared-hardship mechanism for building group loyalty.
• ~92-96%: Nutrition science as a likely-oversimplified field, following on a previous guest's (Laszlo Bassi) claims about under-studied food molecules beyond macronutrients.