Input–output maps are strongly biased towards simple outputs

Dingle, Kamaludin; Camargo, Chico Q.; Louis, Ard A.

doi:10.1038/s41467-018-03101-6

Cited by 72 publications

(229 citation statements)

References 32 publications

Supporting

Mentioning

218

Contrasting

Order By: Relevance

“…This suggests that low evolvability of complex reaction norms is a consequence of the scarcity in the GRN space of the networks that generate complex reaction norms. A similar bias towards simple inputoutput maps has been recently proposed for artificial neural networks [19], and algorithmic theory suggests that it might be a general property of every computable functions [31]. Our contribution is to demonstrate that this bias is intrinsic to cell plasticity, and that it emerges from reasonable model assumptions of gene-expression mechanisms.…”

Section: Discussionsupporting

confidence: 56%

“…a nonlinear and non-additive function of the inputs). As such, Ω is not determined by the number of environmental inputs, but by how complicated is the logic by which they are linked to the cell state [12], [31]. Each logical function is summarised in a so called truth table, a mathematical map which relates each combination of binary environmental factors (EFs) with a given binary output.…”

Section: Resultsmentioning

confidence: 99%

“…Ultimately, the ability of plastic cells to mimic logical functions and to perform proper learning with generalisation is likely to arise from the structural, dynamical, and functional isomorphisms between GRNs and NNs: they both partition the space of possible inputs into a series of states or attractors [13], [37], [42], both are adaptable via incremental improvement principles [15], [41], [46] and both exhibit similar inductive biases [14], [31]. In addition, both exhibit increased performance under similar adaptive pressures [26].…”

Section: Discussionmentioning

confidence: 99%

See 2 more Smart Citations

How to fit in: The learning principles of cell differentiation

Brun‐Usan

Watson

2019

Preprint

View full text Add to dashboard Cite

AbstractCell differentiation in multicellular organisms requires cells to respond to complex combinations of extracellular cues, such as morphogen concentrations. However, most models of phenotypic plasticity assume that the response is a relatively simple function of a single environmental cue. Accordingly, a general theory describing how cells should integrate multi-dimensional signals is lacking.In this work, we propose a novel theoretical framework for understanding the relationships between environmental cues (inputs) and phenotypic responses (outputs) underlying cell plasticity. We describe the relationship between environment and cell phenotype using logical functions, making the evolution of cell plasticity formally equivalent to a simple categorisation learning task. This abstraction allows us to apply principles derived from learning theory to understand the evolution of multi-dimensional plasticity.Our results show that natural selection is capable of discovering adaptive forms of cell plasticity associated with arbitrarily complex logical functions. However, developmental dynamics causes simpler functions to evolve more readily than complex ones. By using conceptual tools derived from learning theory we further show that under some circumstances, the evolution of plasticity enables cells to display appropriate plastic responses to environmental conditions that they have not experienced in their evolutionary past. This is possible when the complexity of the selective environment mirrors the developmental bias favouring the acquisition of simple plasticity functions – an example of the necessary conditions for generalisation in learning systems.These results show non-trivial functional parallelisms between learning in neural networks and the action of natural selection on environmentally sensitive gene regulatory networks. This functional parallelism offers a theoretical framework for the evolution of plastic responses that integrate information from multiple cues, a phenomenon that underpins the evolution of multicellularity and developmental robustness.Author summaryIn organisms composed of many cell types, the differentiation of cells relies on their ability to respond to complex extracellular cues, such as morphogen concentrations, a phenomenon known as cell plasticity. Although cell plasticity plays a crucial role in development and evolution, it is not clear how, and if, cell plasticity can enhance adaptation to a novel environment and/or facilitate robust developmental processes. We argue that available conceptual tools limit our understanding since they only describe simple relationships between the environmental cues (inputs) and the phenotypic responses (outputs) – so called ‘reaction norms’. In this work, we use a new theoretical framework based on logical functions and learning theory that allows us to characterize arbitrarily complex multidimensional reaction norms. By doing this we reveal a strong and previously unnoticed bias towards the acquisition of simple forms of cell plasticity, which increases their ability to adapt to novel environments. Results emerging from this novel approach provide new insights on the evolution of multicellularity and the inherent robustness of the process of development.

show abstract

Section: Discussionsupporting

confidence: 56%

Section: Resultsmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

How to fit in: The learning principles of cell differentiation

Brun‐Usan

Watson

2019

Preprint

View full text Add to dashboard Cite

show abstract

“…This kind of predictive coding is data compression -a mechanism that works to represent a long history of environmental inputs into a compact representation via tuning of internal states. This basic learning and inference capability is the primal origin of "understanding" and intelligence in advanced brains, which can be defined as the process of inferring patterns from raw data and making "maps" of regularities in observations that occupies every Self from the simplest of organisms to the scientist working on extracting deep theory from data (Dingle et al, 2018).…”

Section: A B Cmentioning

confidence: 99%

The Computational Boundary of a “Self”: Developmental Bioelectricity Drives Multicellularity and Scale-Free Cognition

2019

View full text Add to dashboard Cite

All epistemic agents physically consist of parts that must somehow comprise an integrated cognitive self. Biological individuals consist of subunits (organs, cells, and molecular networks) that are themselves complex and competent in their own native contexts. How do coherent biological Individuals result from the activity of smaller sub-agents? To understand the evolution and function of metazoan creatures' bodies and minds, it is essential to conceptually explore the origin of multicellularity and the scaling of the basal cognition of individual cells into a coherent larger organism. In this article, I synthesize ideas in cognitive science, evolutionary biology, and developmental physiology toward a hypothesis about the origin of Individuality: "Scale-Free Cognition." I propose a fundamental definition of an Individual based on the ability to pursue goals at an appropriate level of scale and organization and suggest a formalism for defining and comparing the cognitive capacities of highly diverse types of agents. Any Self is demarcated by a computational surface -the spatio-temporal boundary of events that it can measure, model, and try to affect. This surface sets a functional boundarya cognitive "light cone" which defines the scale and limits of its cognition. I hypothesize that higher level goal-directed activity and agency, resulting in larger cognitive boundaries, evolve from the primal homeostatic drive of living things to reduce stress -the difference between current conditions and life-optimal conditions. The mechanisms of developmental bioelectricity -the ability of all cells to form electrical networks that process informationsuggest a plausible set of gradual evolutionary steps that naturally lead from physiological homeostasis in single cells to memory, prediction, and ultimately complex cognitive agents, via scale-up of the basic drive of infotaxis. Recent data on the molecular mechanisms of pre-neural bioelectricity suggest a model of how increasingly sophisticated cognitive functions emerge smoothly from cell-cell communication used to guide embryogenesis and regeneration. This set of hypotheses provides a novel perspective on numerous phenomena, such as cancer, and makes several unique, testable predictions for interdisciplinary research that have implications not only for evolutionary developmental biology but also for biomedicine and perhaps artificial intelligence and exobiology.

show abstract

“…This is represented by low and high Ω, as defined below (0�Ω�1). As such, Ω is not determined by the number of environmental inputs, but by how complicated is the logic by which they are linked to the cell state [12,44].…”

Section: Measuring the Complexity Of A Multi-dimensional Reaction Normmentioning

confidence: 99%

How to fit in: The learning principles of cell differentiation

2020

View full text Add to dashboard Cite

Cell differentiation in multicellular organisms requires cells to respond to complex combinations of extracellular cues, such as morphogen concentrations. Some models of phenotypic plasticity conceptualise the response as a relatively simple function of a single environmental cues (e.g. a linear function of one cue), which facilitates rigorous analysis. Conversely, more mechanistic models such those implementing GRNs allows for a more general class of response functions but makes analysis more difficult. Therefore, a general theory describing how cells integrate multi-dimensional signals is lacking. In this work, we propose a theoretical framework for understanding the relationships between environmental cues (inputs) and phenotypic responses (outputs) underlying cell plasticity. We describe the relationship between environment and cell phenotype using logical functions, making the evolution of cell plasticity equivalent to a simple categorisation learning task. This abstraction allows us to apply principles derived from learning theory to understand the evolution of multi-dimensional plasticity. Our results show that natural selection is capable of discovering adaptive forms of cell plasticity associated with complex logical functions. However, developmental dynamics cause simpler functions to evolve more readily than complex ones. By using conceptual tools derived from learning theory we show that this developmental bias can be interpreted as a learning bias in the acquisition of plasticity functions. Because of that bias, the evolution of plasticity enables cells, under some circumstances, to display appropriate plastic responses to environmental conditions that they have not experienced in their evolutionary past. This is possible when the selective environment mirrors the bias of the developmental dynamics favouring the acquisition of simple plasticity functions-an example of the necessary conditions for generalisation in learning systems. These results illustrate the functional parallelisms between learning in neural networks and the action of natural selection on environmentally sensitive gene regulatory networks. This offers a theoretical framework for the evolution of plastic responses that integrate information from multiple cues, a phenomenon that underpins the evolution of multicellularity and developmental robustness.

show abstract

Input–output maps are strongly biased towards simple outputs

Cited by 72 publications

References 32 publications

How to fit in: The learning principles of cell differentiation

How to fit in: The learning principles of cell differentiation

The Computational Boundary of a “Self”: Developmental Bioelectricity Drives Multicellularity and Scale-Free Cognition

How to fit in: The learning principles of cell differentiation

Contact Info

Product

Resources

About