Peter Fedichev recently posted a Substack critical of Pedro Magalhaes’ theoretical paper titled “A Brief History of the Hyperfunction Theory of Aging and Future Directions.” Peter’s Substack is pasted below. Unlike Peter, I liked Pedro’s review. I found it illuminating in that it shows some of the history of how one truth-seeking gerontologist struggled to find a model of aging that best fits experimental data. I think Pedro is right about many of his conclusions. Here I add my two cents on this important ongoing debate:   Pedro’s paper is a discussion of the evolution of his conceptual model of the central mechanisms of aging together with some reasoning that led him to eventually lean toward programmatic models. By “programmatic” he means that the genetic program that regulates the pace and complex steps of turning a single cell (the fertilized egg cell) into a human being with trillions of cells extends into the aging process. The idea is that as we age, these processes continue in an unhelpful manner leading to “hyperfunction” (i.e. the progression of development leads to changes that damage the organism in what we call “aging”).   Peter largely objects to Pedro’s model saying it requires “A locus, where the schedule is kept. A timebase, so the schedule has a tempo. And a channel, carrying instructions from the locus to the targets.” He then describes mathematic models that do not require the above.    I must say, as much as I value mathematics and physics, when applied to biology and medicine, mathematical models to try to predict yet-undiscovered molecular processes within actual living cells has a mixed track record. For example… Cell senescence (the Hayflick phenomenon) used to be considered as entropy on display, complex deterioration that we would never understand or intervene in. Sohl, D. A., & Witten, M. (1995) wrote a paper titled Mathematical modeling of cellular aging and senescence: A general model of the Hayflick limit. Journal of Theoretical Biology, 175(2), 117–128. They analyzed cell population dynamics using systems of differential equations and stochastic matrices to account for the heterogeneous decay observed in human diploid fibroblasts. They argued that senescence reflected a high degree of systemic biological complexity that could not be explained by simple uniform mechanisms.   Based on empirical data at Geron we found the telomerase gene and showed that it alone could immortalize cells. The mathematical modeling, even when performed at about the same time, was wildly off-track.   Similarly, it was widely accepted prior to the cloning of Dolly the sheep that the reprogramming of a mammalian genome by SCNT wasn’t possible. Indeed McGrath and Solter stated in 1984 “the cloning of mammals by nuclear transfer is biologically impossible.” We now know SCNT can even can reverse cell aging (our Science paper in 2000).   Back to Peter’s statement that programmatic theories require a timebase, tempo, and channel to actuated the clock, I think he is exactly right. But I would suggest such pathways exist and regulate developmental timing (heterochrony). Example in point, the increased lifespan of humans vs non-human primates widely recognized to be due to heterochrony (neoteny).   Bottom line, IMO, science best progresses by proposing a hypothesis, identifying experiments to test and potentially falsify that hypothesis, then modifying the hypothesis to fit those new empirical observations. Experiments to test the role of molecular pathways involved in heterochrony and development in aging are currently underway and before the end of this decade we will have a much clearer understanding which model is most accurate. My prediction is that the heterochrony pathways will tie together observations about dietary restriction, mTOR, and regeneration, etc.. The development of these models will develop and function (or hyperfunction) and time will tell…   xitter.cf/fedichev/status/209010… @jpsenescence @aubreydegrey @fedichev