Biology subjects

Barnes, C. P.

Publications and source records attributed to Barnes, C. P..

At least 19 recordsLinked to original sources

bioRxiv2026

Nonlinear regime transitions enable reservoir computation in activator-inhibitor cellular automata

Activator-inhibitor systems are usually studied as pattern-forming media, but the same local nonlinear interactions can also shape how information is stored and separated over time. Here we first introduce a reaction-diffusion-inspired cellular automaton as a tunable nonlinear medium with two activation families, a continuous sigmoid and a logistic-step relaxer. In each case, a single family-specific parameter changes the nonlinear response or relaxation dynamics while leaving the neighbourhood wiring fixed. This provides a controlled way to move the activator-inhibitor system between collapsed, structured, saturated, and overshooting regimes, and to relate these regimes to state-level diversity and pattern compressibility. We then use reservoir computing to test whether these tuned pattern-forming regimes support fading memory, input separability, and downstream readout learning. In this role, reservoir computing is not treated as an architecture to be optimised, but as a diagnostic of the computational properties of the untrained medium. Across parameter sweeps, stronger memory and separability are concentrated near transition regions rather than distributed uniformly across parameter space. These results identify activator-inhibitor cellular automata as interpretable unconventional reservoir substrates in which effective gain, threshold, and relaxation parameters tune the balance between pattern formation, memory, and separability. More broadly, they support the view that tissue-like or physical pattern-forming media may, in principle, shift between patterning and information-processing regimes by modulating local nonlinear dynamics.

systems biology↗

bioRxiv2026

PlasmidLM: A Promptable DNA Language Model via Verifiable-Reward Post-Training

Generative DNA models are typically next-token completers: they extend a sequence but offer no native interface for telling the model what to make. PlasmidLM is a promptable DNA language model for plasmids. A designer supplies a human-readable component specification, for example a high-copy E. coli vector with kanamycin resistance and an EGFP reporter, and the model generates the corresponding multi-kilobase construct in a single autoregressive pass. Prompts are unordered sets of named-part tokens at the granularity of biological shorthand, not learned latent codes or rigid grammars. We evaluate outputs along two axes: a sequence is viable if structurally plausible as a plasmid, and faithful if its detected components match the prompt. Their conjunction is the useful-plasmid rate, the primary metric we report. On a held-out 1,000-prompt benchmark, the post-trained model achieves a useful-plasmid rate of 48.5% at single-shot decoding and 89.7% under best-of-4 sampling. Verifiable-reward post-training with GRPO against a 660-entry sequence motif registry improves the useful-plasmid rate across all sampling budgets. We release the 19.3M-parameter model, evaluation suite, and a paired benchmark of prompt-sequence pairs.

bioengineering↗

bioRxiv2026

Emergent Biological Realism in RL-Trained DNA Language Models

We compare the efficacy and distributional effects of supervised fine-tuning (SFT) and reinforcement learning (RL) post-training for PlasmidGPT, a foundation model for whole-plasmid generation, using Group Relative Policy Optimization (GRPO) for the RL model. Using a biologically motivated reward function encoding functional annotations, length constraints, and repeat penalties, the RL model achieves a 71.6% quality-control pass rate across 8 prompts on 4,000 sequences, compared to 4.3% for the pretrained baseline and 11.0% for SFT. A five-model reward ablation identifies the cassette arrangement bonus, which rewards correct promoter [->] CDS [->] terminator ordering, as the critical reward component. Rejection-sampling baselines indicate that the gain is not recovered by sampling more heavily from the base model. Beyond directly optimized features, RL-generated sequences converge toward real plasmid distributions in 3-mer composition and minimum free energy density, neither of which is directly optimized by the reward function. Minimum free energy density independently converges to the real-plasmid regime under both SFT and RL despite these being parallel post-training paths. On a small curated hold-out set, RL improves continuation log-likelihood over the pretrained baseline on all 29 held-out sequences (mean {Delta} = +0.83 nats).

bioinformatics↗

bioRxiv2025

Generative design and construction of functional plasmids with a DNA language model

DNA language models offer a new paradigm for sequence design, yet their ability to generate functional genomic sequences remains underexplored. Plasmids act as a good testbed for evaluating DNA language model generation potential due to their simplicity and ease of construction. Here, we develop an end-to-end pipeline for generative design of Escherichia coli plasmid backbones, from large-scale data curation through fine-tuning, sampling, bioinformatic assessment, and candidate selection. A curated plasmid library was assembled from PlasmidScope and Addgene, and PlasmidGPT, a GPT-2-style DNA model, was fine-tuned on these corpora using circular-aware batching and random crops. Generations (1,000 per model) were produced under two prompting strategies: a minimal ATG seed to expose default tendencies, and a GFP cassette to enforce functional context. From 1000 generated synthetic plasmids, 16 candidates survived strict filtering and these were prioritised for wet-lab validation. Three shortlisted plasmids were synthesised and found to be functional, supporting growth, antibiotic resistance, and GFP expression in E. coli. These represent, to our knowledge, the first full AI-generated plasmids to be synthesised and validated in vivo. This work demonstrates that curated fine-tuning and prompt-aware generation enable DNA language models to progress from raw sequence sampling to experimentally testable plasmid designs. The approach offers a foundation for extending DNA design optimisation beyond E. coli, toward broader applications across engineering biology.

synthetic biology↗

bioRxiv2025

Quantitative measurement of phenotype dynamics during cancer drug resistance evolution using genetic barcoding

Cancer treatment frequently fails due to the evolution of drug-resistant cell phenotypes driven by genetic or non-genetic changes. The origin, timing, and rate of spread of these adaptations are critical for understanding drug resistance mechanisms but remain challenging to observe directly. We present a mathematical framework to infer drug resistance dynamics from genetic lineage tracing and population size data without direct measurement of resistance phenotypes. Simulation experiments demonstrate that the framework accurately recovers ground-truth evolutionary dynamics. Experimental evolution to 5-Fu chemotherapy in colorectal cancer cell lines SW620 and HCT116 validates the framework. In SW620 cells, a stable pre-existing resistant subpopulation was inferred, whereas in HCT116 cells, resistance emerged through phenotypic switching into a slow-growing resistant state with stochastic progression to full resistance. Functional assays, including scRNA-seq and scDNA-seq, validate these distinct evolutionary routes. This framework facilitates rapid characterisation of resistance mechanisms across diverse experimental settings.

cancer biology↗

bioRxiv2024

Engaging cancer patients on their attitudes towards microbiome engineering technologies

Microbiome engineering aims to develop engineered live biotherapeutic products to diagnose and treat human disease. One of the most active areas is in the engineering of modified bacteria that act as cancer therapeutics, with many products currently in clinical trials. With the emergence of any novel technologies, it is important to consult the public and end users throughout their development to address any concerns that may arise from their use. This is particularly true for therapeutics, for which it is vital that both clinicians and patients have confidence in the treatments that are available to them. Here, we surveyed a cohort of cancer patients and their relatives to gauge their knowledge of modern microbiome engineering techniques. We focus predominately on the use of bacterial eLBPs to treat cancer. Overall, most participants indicated that they would be comfortable taking cancer treatment options based on microbiome engineering technologies.

scientific communication and education↗

bioRxiv2024

Quantifying the intricacies of biological pattern formation: A new perspective through complexity measures

Nonlinear pattern-forming systems can generate spatial structure through many mechanisms, but comparing the patterns that are actually realized after the dynamics unfold remains difficult. Linear stability analysis identifies when spatial modes can grow and which length scales are favored near onset, but it does not directly measure the state diversity, structural organization, or dynamical route of the resulting patterns. Here we present a comparative operational framework, mechanism-agnostic under a fixed measurement protocol, based on two coordinates: the Diversity of Number of States (DNOS), an occupied-state measure of effective alphabet breadth, and the Diversity of Pattern Complexity (DPC), a lossless-compression measure of structural complexity. We apply the framework to a linear reaction--diffusion benchmark, the Gray--Scott and FitzHugh--Nagumo models, toggle-switch gene-regulatory systems with and without self-activation, and a Notch--Delta--EGF model of Drosophila neurogenesis. Across these systems, DNOS--DPC landscapes identify realized pattern regimes not determined by linear onset analysis alone. Robustness checks against bin count, entropy-based DNOS diagnostics, compression settings, image noise, clustering method, and graph, spectral, and wavelet descriptors support the stability of the comparative summaries under the stated protocol. In the Drosophila model, endpoint, field-resolved, multichannel, and time-resolved analyses distinguish pathway-level Notch/Delta signaling collapse from partial downstream proneural-output persistence, and suggest an EGF-dependent entry/progression bias into organized proneural-wave states. The framework provides reproducible coordinates for comparing realized pattern diversity, structural complexity, and developmental trajectories across nonlinear reaction--diffusion and regulatory systems.

developmental biology↗

bioRxiv2024

Designing minimal E. coli genomes using variational autoencoders

Designing minimal bacterial genomes remains a key challenge in synthetic biology. There is currently a lack of efficient tools for the rapid generation of streamlined bacterial genomes, limiting research in this area. Here, using a pangenome dataset for Escherichia coli, we show that variational autoencoders with modified loss functions can successfully create minimised genomes retaining the essential genes identified in the literature. We then sampled new genomes from our fitted model and performed computational validation using an E. coli whole-cell model. We found 6 out of 100 of the sampled genomes were viable in the computer model. These underwent a minimization routine starting from the MG1655 genome giving rise to six new minimal genomes with around a 40 % reduction in size. This study proposes a rapid, machine learning-based approach for bacterial sequence generation, that could accelerate the genomic design process.

synthetic biology↗

bioRxiv2024

Bacterial microcompartment utilisation in the human commensal Escherichia coli Nissle 1917

Bacterial microcompartments (BMCs) are self-assembled protein structures often utilised by bacteria as a modular metabolic unit, enabling the catalysis and utilisation of less common carbon and nitrogen sources within a self-contained compartment. The ethanolamine (EA) utilisation (eut) BMC has been widely demonstrated in enteropathogens, such as Salmonella enterica, and current research is exploring its activity in the commensal species that populate the human gut. Escherichia coli Nissle 1917 (EcN) is a strong coloniser and probiotic in gut microbial communities, and has been used extensively for microbiome engineering. In this study, the utilisation of ethanolamine as a sole carbon source and the formation of the eut BMC in EcN were demonstrated through growth assays and visualisation with transmission electron microscopy. Subsequently, flux balance analysis was used to further investigate the metabolic activity of this pathway. It was found that not only is the utilisation of the eut BMC for the degradation of EA as a carbon source in EcN comparable to that of Salmonella enterica, but also that ammonium is released into solution as a byproduct in EcN but not in S. enterica. Control of EA-dependent growth was demonstrated using different concentrations of the operon inducer, vitamin-B12. We show that vitamin B12-dependent EA utilisation as the sole carbon source enables growth in EcN, and demonstrate the concurrent formation of the BMC shell and inducible control of the eut operon.

microbiology↗

bioRxiv2024

Explainable deep learning on 7500 whole genomes elucidates cancer-specific patterns of chromosomal instability

Chromosomal instability leads to an increased rate of chromosome or part-chromosome copy gains or losses. Referred to as chromosome copy aberrations (CNAs), these mutations are highly prevalent in cancer cells and contribute to abnormal genome structure, genetic diversity within a tumour, drug resistance, and evolution to metastatic disease. Within each cancer type, it remains unclear how CNAs define each particular malignant phenotype, yet some key events have been characterized. Here, we present a novel deep learning method, DeepCNA, and apply it as part of an exploration of CNAs present across 7,500 whole genomes and 13 cancer types. DeepCNA is an explainable AI approach, that reveals both established and novel loci that contribute to cancer type. It can be used as a diagnostic tool in situations such as cancer of unknown primary site, and it can discover CNAs that are prognostic. It also has several applications for researchers, including drug target and biomarker discovery.

cancer biology↗

bioRxiv2024

Cell-cycle dependent DNA repair and replication unifies patterns of chromosome instability

Chromosomal instability (CIN) is pervasive in human tumours and often leads to structural or numerical chromosomal aberrations. Somatic structural variants (SVs) are intimately related to copy number alterations but the two types of variant are often studied independently. In addition, despite numerous studies on detecting various SV patterns, there are still no general quantitative models of SV generation. To address this issue, we develop a computational cell-cycle model for the generation of SVs from end-joining repair and replication after double strand break formation. Our model provides quantitative information on the relationship between breakage fusion bridge cycle, chromothripsis, seismic amplification, and extra-chromosomal circular DNA. Given single-cell whole-genome sequencing data, the model also allows us to infer important parameters in SV generation with Bayesian inference. Our quantitative framework unifies disparate genomic patterns resulted from CIN, provides a null mutational model for SV, and reveals new insights into the impact of genome rearrangement on tumour evolution.

genomics↗

bioRxiv2023

An antimicrobial peptide expression platform for targeting pathogenic bacterial species

Bacteriocins are antimicrobial peptides that are naturally produced by many bacteria. They hold great potential in the fight against antibiotic resistant bacteria, including ESKAPE pathogens. However, they often have low stability in vivo and therefore, may not be effective when administered orally. Engineered live biotherapeutic products (eLBPs) that secrete bacteriocins can be created to deliver bacteriocins directly at the site of infection. Here we develop a modular bacteriocin secretion platform that can be used to express and secrete multiple bacteriocins from non-pathogenic Escherichia coli host strains. As a proof of concept we create Enterocin A and Enterocin B secreting strains that show strong antimicrobial activity against Enterococcus faecalis and Enterococcus faecium, and characterise this activity in both solid culture and liquid co-culture. We then develop a Lokta-Volterra model that can be used to capture the interactions of these competitor strains. We show that simultaneous exposure to EntA and EntB can delay the point of Enterococcus growth. Our system has the potential to be used as an eLBP to secrete additional bacteriocins for the targeted killing of other pathogenic bacteria.

synthetic biology↗

bioRxiv2023

Emergent digital bio-computation through spatial diffusion and engineered bacteria

Building computationally capable biological systems has long been an aim of synthetic biology. The potential utility of biocomputing devices ranges from biosafety and environmental applications to diagnosis and personalised medicine. Here we present work for the design of bacterial computers which use spatial patterning to process information. Our computers are composed of bacterial colonies which, inspired by patterning in embryo development, receive information in the form of diffusible morphogen-like signals. A computation is encoded by the physical locations of the input sources of morphogen and the output receiver colonies. We demonstrate, mathematically and with engineered Escherichia coli, the simple digital logic capability of single bacterial colonies and show how additional colonies are required to build complex functions. Inspired by electronic design automation, an algorithm for designing optimal spatial circuits computing two-level digital functions is presented. This enhances the capability of our system to complex digital functions without increasing the biological complexity. We extend our experimental system to incorporate biosensing colonies as morphogen sources, demonstrating how a diagnostic device might be constructed. Our approach will open up new ways to perform biological computation, with applications in bioengineering, biomaterials and biosensing. Ultimately, these computational bacterial communities will help us explore information processing in natural biological systems.

synthetic biology↗

bioRxiv2022

Automated design of gene circuits with optimal mushroom-bifurcation behaviour

Recent advances in synthetic biology are enabling exciting technologies, including the next generation of biosensors, the rational design of cell memory, modulated synthetic cell differentiation and generic multi-functional bio-circuits. These novel applications require the design of gene circuits leading to sophisticated behaviours and functionalities. At the same time, designs need to be kept minimal to avoid compromising cell viability. Bifurcation theory of dynamical systems provides powerful tools to address complex nonlinear dynamics and multifunctionality, linking model topology and kinetic parameters with circuit behaviour. However, the challenge of incorporating bifurcation analysis to automated design has not been accomplished so far. In this work we present an optimisation-based method for the automated forward design of synthetic gene circuits with specified bifurcation diagrams, allowing us to find minimal topologies optimizing the required functionalities and taking into account additional requirements and/or context specifications. We apply the method to design of gene circuits exhibiting the so called mushroom bifurcation, a relatively unexplored multi-functional behaviour of particular relevance for developmental biology. Using the results of the optimisation analysis we explore the capabilities of the resulting circuits for possible applications in advanced biosensors, memory devices, and synthetic cell differentiation.

systems biology↗

bioRxiv2022

Deep Reinforcement Learning for Optimal Experimental Design in Biology

The field of optimal experimental design uses mathematical techniques to determine experiments that are maximally informative from a given experimental setup. Here we apply a technique from artificial intelligence--reinforcement learning--to the optimal experimental design task of maximizing confidence in estimates of model parameter values. We show that a reinforcement learning approach performs favourably in comparison with a one-step ahead optimisation algorithm and a model predictive controller for the inference of bacterial growth parameters in a simulated chemostat. Further, we demonstrate the ability of reinforcement learning to train over a distribution of parameters, indicating that this approach is robust to parametric uncertainty. 1 Author summaryBiological systems are often complex and typically exhibit non-linear behaviour, making accurate model parametrisation difficult. Optimal experimental design tools help address this problem by identifying experiments that are predicted to provide maximally accurate parameter estimates. In this work we use reinforcement learning, an artificial intelligence method, to determine such experiments. Our simulation studies show that this approach allows uncertainty in model parameterisation to be directly incorporated into the search for optimal experiments, opening a practical avenue for training an experimental controller without confident knowledge of the systems parameter values. We present this method as complementary to existing optimisation approaches and we anticipate that artificial intelligence has a fundamental role to play in the future of optimal experimental design.

synthetic biology↗

bioRxiv2022

Experimental evolution in TP53 deficient gastric organoids recapitulates tumorigenesis

The earliest events during human tumor initiation, while poorly characterized, may hold clues to malignancy detection and prevention1. Here we model occult pre-neoplasia by bi-allelically inactivating TP53, a common early event in gastric cancer, in human gastric organoids. Causal relationships between this initiating genetic lesion and resulting phenotypes were established using experimental evolution in multiple clonally derived cultures over two years. TP53 loss elicited progressive aneuploidy, including copy number alterations and structural variants prevalent in gastric cancers, with evident preferred orders. Longitudinal single cell sequencing of TP53 deficient gastric organoids similarly indicates progression towards malignant transcriptional programs. Moreover, high-throughput lineage tracing with expressed cellular barcodes demonstrates reproducible dynamics whereby initially rare subclones with shared transcriptional programs repeatedly attain clonal dominance. This powerful platform for experimental evolution exposes stringent selection, clonal interference, and a striking degree of phenotypic convergence in pre-malignant epithelial organoids. These data imply predictability in the earliest stages of tumorigenesis and reveal evolutionary constraints and barriers to malignant transformation with implications for earlier detection and interception of aggressive, genome instable tumors.

cancer biology↗

bioRxiv2022

CNETML: Maximum likelihood inference of phylogeny from copy number profiles of spatio-temporal samples

Phylogenetic trees based on copy number alterations (CNAs) for multi-region samples of a single cancer patient are helpful to understand the spatio-temporal evolution of cancers, especially in tumours driven by chromosomal instability. Due to the high cost of deep sequencing data, low-coverage data are more accessible in practice, which only allow the calling of (relative) total copy numbers due to the lower resolution. However, methods to reconstruct sample phylogenies from CNAs often use allele-specific copy numbers and those using total copy number are mostly distance matrix or maximum parsimony methods which do not handle temporal data or estimate mutation rates. In this work, we developed a new maximum likelihood method based on a novel evolutionary model of CNAs, CNETML, to infer phylogenies from spatio-temporal samples taken within a single patient. CNETML is the first program to jointly infer the tree topology, node ages, and mutation rates from total copy numbers when samples were taken at different time points. Our extensive simulations suggest CNETML performed well even on relative copy numbers with subclonal whole genome doubling events and under slight violation of model assumptions. The application of CNETML to real data from Barretts esophagus patients also generated consistent results with previous discoveries and novel early CNAs for further investigations.

bioinformatics↗

bioRxiv2021

Chaos in small microbial communities

Predictability is a fundamental requirement in biological engineering. As we move to building coordinated multicellular systems, the potential for such systems to display chaotic behaviour becomes a concern. Therefore understanding which systems show chaos is an important design consideration. We developed a methodology to explore the potential for chaotic dynamics in small microbial communities governed by resource competition, intercellular communication and competitive bacteriocin interactions. We show that we can expect to find chaotic states in relatively small synthetic microbial systems, understand the governing dynamics and provide insights into how to control such systems. This work is the first to query the existence of chaotic behaviour in synthetic microbial communities and has important ramifications for the fields of biotechnology, bioprocessing and synthetic biology.

microbiology↗