In the last 4 posts of this study I’ve built up an argument showing the need for a definition of biological aging that consolidates existing consensus knowledge in the field but also flexible enough to incorporate new knowledge within that current paradigm.
This is still not the final formulation, but this is what I have right now:
‘Biological aging is agings underneath, the result of multiple, separate, diverse, interconnected, but malleable processes, eventually compromising normal functions of the organism at different rates and at all (organisational, spatiotemporal) levels.’
What are some open problems left with this definition?
I see two main problems, one is related to causality, the other related to what I call the ‘computability’ of the definition.
The ‘causality’ problem has 2 parts: a conceptual and a mathematical. The conceptual can be stated the following way: The definition is pointing towards different and identified molecular/cellular processes that are proposed to be the main causes of overall biological aging. These are detectable/measurable with state of the art biological methods and tools. Since the recursive definition in its current form is heavily relying on and built out of The Hallmarks Framework it’s important to understand how causality is to be understood. Particularly, need to understand what forms necessary and what forms sufficient causes within these processes, which are the ones that other process cannot do without and which are the ones that trigger further events? Now the internal structure of the Hallmark processes, the one that we have used during recursion, primary base cases, antagonistic and integrative hallmarks suggest an interpretation in terms of causality but this needs to be tested further.
Here, the mathematical problem comes into question and with that the mathematical representation of the causal structure of biological aging(s). What is needed here is to use the causal calculus and causal inference methods, built upon Bayesian networks, championed by Judea Pearl, see The Book of Why, to give the Hallmarks framework a proper mathematical formulation. And that can power our definition. I started to work towards that understanding in my causation and aging series.
This leads us to the ‘computability’ aspect, strictly linking back to causality modelling. By ‘computability’ of the definition I mean it’s usefulness in guiding mathematical, quantitative formulations and tests to invent reliable biomarkers and bio-patterns of aging and inform designing interventions to slow, stop or reset aging processes. I argued before for the usefulness of this definition and I think it’s good at exhibiting a ‘divide and conquer’ approach. But some math, some simulations, modelling and some actual code is needed here. Causal networks might suffice for the task but there are a lot of other options, several different flavours of math, probability, statistics and machine learning to consider.
These two questions, causality and computability are the ones that stand out as the most relevant to give this definition a substance.