Researchers deny the validity of the “set theory of everything” hypothesis

Credit: Pixabay/CC0 Public Domain
Three new papers refute the ensemble theory of molecular complexity, which has been billed as a new “theory of everything.”
First publicly proposed in 2017, assembly theory is a hypothesis about the measurable complexity of molecules that claims to characterize life, explain natural selection and evolution, and even redefine our understanding of time, matter, life, and the universe.
However, researchers led by Dr Hector Zenir from the School of Biomedical Engineering and Imaging Sciences (BMEIS), together with colleagues from King Abdullah University of Science and Technology (KAUST) and Karolinska Institutet in Sweden, have now succeeded in demonstrating in a paper published in the journal npj Systems Biology that the same result can be achieved using traditional statistical and compression algorithms.
A second paper recently published by PLoS Complex Systems mathematically proves that assembly theory is equivalent to Shannon entropy and is therefore not a novel approach to any of these applications, but rather an implementation of a well-known and common compression algorithm used behind image encoding formats such as ZIP compression and PNG.
The third paper, “Assembly theory reduced to Shannon entropy and made redundant by a simple statistical algorithm,” is available on the arXiv preprint server.
“Our study demonstrates that a core element of assembly theory – the assembly index, which determines the ‘viability’ of an object by the number of exact copies it has – is not the way it should be and that its conclusions are flawed,” says Dr Hector Zenil.
“Applying traditional compression algorithms to molecular and chemical data produced the same validated results as in assembly theory. This means that assembly theory is not a new framework and is indistinguishable from existing complexity measures. However, the original authors did not test other algorithms.”
“Even though some vegetables and plants, such as onions and ferns, have a large number of gene copies and genomes that are up to 50 times longer, it is difficult to argue that they are more complex and alive than humans, as assembly theories based on such one-dimensional markers would suggest,” says Professor Jesper Tegner.
“What truly defines life is not simply the length of its genes or the number of its components, but the complex relationships it has with its environment, the agency it exhibits, and the resilience it has to maintain its essential properties.”
“Our analysis reveals the limitations of assembly theory’s numerical indicators, which seek to define the ‘survival’ of life and vital properties. What really surprised me is that the crucial role of dynamic interactions in understanding the complexity of life is ignored. Even more alarming is the decision to propose a fixed life detection threshold without any basis,” said Dr. Narcisse A. Kiani.
“The real breakthroughs will come from using the tools we’ve already developed to build on established knowledge, rather than rehash what we already know, and integrate seemingly diverse theories to uncover the complex, multidimensional dynamics that shape life.”
Characterizing life is difficult and still an unsolved problem, but it has been studied from many angles, from Gregor Mendel’s modular units to Erwin Schrödinger’s thermodynamics, Claude Shannon’s statistical entropy, and Gregory Chaitin’s algorithmic information.
Today, with all this knowledge and even more from complexity science and systems biology, we know that one of the key aspects of life is its open-endedness, that is, the fact that the agency of life is not limited to regular actions and repetitions in adaptation and relationship to the environment.
Fields such as Algorithmic Information Dynamics (AID), led by Dr. Hector Zenil and his collaborators, are shedding light on how to find causal models of natural phenomena and mechanistic explanations of processes in living systems.
AID builds on current knowledge in information theory and causal inference to build bridges between these fundamental disciplines used today to understand the world.
The method behind AID already counts exact copies of modules, the most obvious first step, and something that Dr Zenir reported prior to assembly theory that allowed him to separate organic from non-organic compounds as a function of molecular length.
Further information: Abicumaran Uthamacumaran et al. “On assembly theory methods and notable limitations for classification of molecular biosignatures.” npj Systems Biology and Applications (2024). DOI: 10.1038/s41540-024-00403-y
Felipe S. Abrahão et al., “Assembly theory is an approximation of the complexity of algorithms based on LZ compression that cannot account for selection or evolution,” PLOS Complex Systems (2024). DOI: 10.1371/journal.pcsy.0000014
Luan Ozelim et al. “Reducing assembly theory to Shannon entropy and redundancy with a simple statistical algorithm.” arXiv (2024). DOI: 10.48550/arxiv.2408.15108
Journal information: arXiv
Courtesy of King’s College London
Citation: Researchers refute validity of “assembly theory of everything” hypothesis (September 24, 2024) Retrieved September 24, 2024 from https://phys.org/news/2024-09-refute-validity-theory-hypothesis.html
This document is subject to copyright. It may not be reproduced without written permission, except for fair dealing for the purposes of personal study or research. The content is provided for informational purposes only.