Michael G. Levin

The paper introduces topoiesis, a mechanism where diffusion and geometry drive collective morphogenesis and survival in Xenopus embryos under stress.

This study identifies cholinergic and adrenergic neural signals that activate human intestinal L-cells, triggering hormone secretion and calcium waves.

The paper presents a computational framework to identify interventions that compensate for ion channel mutations by modulating other conductances, demonstrating feasibility through simulations and a differentiable Hodgkin-Huxley model.

The paper introduces MOMbot, a robot scientist that autonomously performs multimodal physical interventions on biological systems to discover how inputs affect cellular and organismal outcomes.

The paper shows that tumor cells use membrane voltage and connexin expression to control communication with neighboring cells, enhancing growth and metastasis.

This study shows that bacterial collectives can encode information about nearby multicellular systems through spatial patterns and machine learning.

The paper introduces a platform called the Cell Trainer for automated training of mammalian cells using timed drug pulses and fluorescence imaging, demonstrating sensitization-like dynamics and feedback control in single-cell responses.

Planarians trained to associate light with food kept their light preference after decapitation and 14-day head regeneration (7.65 min after vs 6.87 min before, out of 10; 10 worms).

The paper investigates transcriptional responses of planarians to BaCl2 stress, showing how regeneration history and social environment influence gene expression.

This study shows that ionic exposure history affects inner nuclear membrane voltage and chromatin texture in NRK cells, with responses dependent on exposure order and chromatin state.

This study shows that purinergic signaling disrupts multi-cellular coordination in basal Xenobots, altering information flow and integration.

This study identifies 243 genome-wide significant loci for hernia subtypes, reveals shared and distinct genetic factors, and identifies potential therapeutic targets.

The paper develops a metabolic model to test if adaptive enzyme regulation enables improved recovery after repeated perturbations and how environmental variability affects metabolic plasticity.

The paper investigates ephaptic-axonal interactions in developing cortical tissue and finds radial biases in neural activity patterns during self-organization.

The paper identifies complex information architectures in Xenopus laevis organoids, suggesting they exhibit integrated, complex systems despite being non-neural.

The paper introduces an AI method to erase memories in gene regulatory networks without altering their structure, with implications for biomedical applications and evolutionary theory.

This study shows TNBC cells activate sensory neurons via TRPV1, driving nerve outgrowth and tumor progression, with TRPV1 knockout reducing tumor growth and metastasis.

The paper uses functional connectivity networks to analyze calcium dynamics in non-neural tissues, revealing non-random structural and informational patterns post-injury.

The paper characterizes self-organizing neural networks in Xenopus-derived biobots, showing they exhibit complex behaviors, active neurons, and distinct transcriptomic profiles compared to non-neuronal constructs.

This paper demonstrates that spontaneous adaptive organisation can arise from the interaction of physical optimisation and learning in dynamical systems, without natural selection.

This paper evaluates plasma biomarkers for causal association with peripheral artery disease using Mendelian randomization, identifying 19 biomarkers with significant genetic evidence for causality.

The paper uses computational models to show how phase synchronisation of microscale agents leads to emergent higher-level agency through increased problem-solving competency.

The paper develops a bioelectric-genetic dynamical model to understand how spatial voltage patterns influence gene expression and brain development in Xenopus embryos, revealing mechanisms of information integration and plasticity.

The paper analyzes regulatory nonlinearity in biological networks using Boolean models and finds that biological regulation is generally less nonlinear than expected, with cancer networks showing higher and more variable nonlinearity.

This paper identifies vorinostat as a potential repurposed therapy for Rett syndrome by combining computational network analysis with CRISPR-edited Xenopus models, showing efficacy across multiple organ systems.

This study shows that ERK inhibition delays regeneration in headless planarians, but headless animals can later remodel to regain a single-headed morphology through Wnt/β-catenin signaling.

This paper introduces a computational framework for damage detection and regeneration in multicellular systems, demonstrating its ability to regenerate tissues and stem cells in simulated organisms.

The paper analyzes protein interactomes across 1800 species to show that biological networks develop macroscales associated with lower noise and higher resilience, particularly in eukaryotes.

The paper identifies vwde expression as a conserved feature of blastemas involved in appendage regeneration across species and shows its functional requirement in regeneration.

This study investigates how HDAC activity epigenetically modulates myeloid cell behavior and regenerative ability in Xenopus tadpoles during tail regeneration.