Biology subjects

Boden, M.

Publications and source records attributed to Boden, M..

At least 19 recordsLinked to original sources

bioRxiv2026

Reframing enzyme function prediction as conditional generation

Enzymes frequently exhibit promiscuous activity beyond their native roles, providing starting-points for new functions. Finding these promiscuous enzymes, especially for non-native chemical transformations, is challenging but highly valuable, as they promise novel, sustainable solutions for chemistry and biotechnology. However, current machine learning methods are poorly suited to discovering unseen chemistry as they often frame function prediction as closed set classification or a retrieval task. Here, we present Fluxion, a generative deep learning framework that learns enzymatic catalysis by modeling dynamic electron flow trajectories across the enzyme's catalytic residues. By combining both synthetic chemistry and biochemical datasets with protein language model representations, Fluxion generates multi-step electron-flow trajectories analogous to the arrow-pushing representations used to describe enzyme reaction mechanisms. Generation is conditioned on enzyme context, including the enzyme sequence, catalytic residues, substrates, and cofactors. We show that this conditioning allows Fluxion to learn enzyme-dependent regioselectivity across cytochrome P450 enzymes with different sequences shifting the predicted reaction sites for the same substrate. We then demonstrate that Fluxion's embeddings are useful for downstream tasks, such as specificity prediction on two experimental datasets, with and without finetuning. Finally, we show that Fluxion has the potential to transfer synthetic chemical logic to biology; it can generate the observed non-native product from real-world non-native directed evolution screens. Our results establish a proof of concept that generative modeling through mechanistic representations of enzymes can shift enzyme function prediction beyond static database retrieval and closed set classification to function generation. This conceptual framework provides a stepping stone towards an in silico generative method to discover non-native biocatalysts.

biochemistry↗

bioRxiv2026

TreeGazer: Prospecting Protein Sequence-Function Landscapes via Phylogenetic Structure

Building diverse and informative protein sequence datasets is critical for understanding how function varies across sequence space. Because only a small fraction of sequences in a dataset can typically be experimentally characterised, strategies for selecting what sequences to characterise should maximise the information gained from each experiment. Here, we present TreeGazer, a phylogeny-informed framework that combines Bayesian optimisation with the topology of a tree to guide sequence selection. TreeGazer balances exploitation of sequences predicted to exhibit favourable properties against exploration of regions higher model uncertainty. Unlike existing approaches that apply Bayesian optimisation for sequence selection, TreeGazer does not rely on black-box models and instead uses latent representations of property distributions that are directly tied to phylogenetic structure. Modelling properties in this way enables biologically interpretable predictions and uncertainty estimates. Across two simulated selection campaigns, TreeGazer consistently selected sequences that produced datasets more representative of the underlying property distribution than alternative strategies that used protein language models. TreeGazer also performed effectively in low-data settings, where tree-guided selection enabled accurate identification of functional transitions across clades. TreeGazer can be run on conventional laptop computers while still providing equivalent or superior performance to embedding-based approaches. These results demonstrate that phylogenetic structure is a powerful and underutilised prior for guiding informative sequence selection.

bioinformatics↗

bioRxiv2026

Systematic Regional Bias is Widespread in ChIP-seq

Robust and reproducible results are essential for confident scientific analysis. We demonstrate that transcription factor (TF) Chromatin Immunoprecipitation coupled with sequencing (ChIP-seq) suffers from systematic bias that may threaten its reproducibility: 80% of 200+ condition-matched, dual-replicate experiments in ENCODE contain genomic regions of systematic bias. We observe this regional bias even between replicates produced within the same experiment, resulting in thousands of unreplicated peaks, which often contain valuable biological data. We provide evidence that regional bias may lead to qualitative differences in TF biology inferred by different experiments; we discovered eight TFs with binding activity in compact chromatin that was identified by one experiment, yet systematically absent from others. To mitigate the effects of bias, we derive simple but effective metrics to quantify the quality of data within biased regions and demonstrate that they can be used for the robust integration of data from multiple experiments.

bioinformatics↗

bioRxiv2026

Combinatorial epigenomic patterns define regulatory programs underlying disease heterogeneity

Complex diseases exhibit substantial variation in clinical presentation and outcome despite shared diagnoses. Current genetic models typically represent inherited risk as a single additive liability, obscuring the diverse biological mechanisms through which variants influence disease. Here, we show that disease-associated variants are organized into recurrent regulatory programs that reveal latent disease mechanisms. Using genome-scale epigenomic maps across human tissues and cell states, we identify regulatory programs that partition disease-associated variants without phenotypic or disease-specific priors. Variants assigned to different programs exert distinct biological effects that translate into divergent clinical outcomes. In type 2 diabetes, these programs reveal previously unrecognized disease subtypes with opposing cardiometabolic profiles that stratify future risk of myocardial infarction and non-alcoholic fatty liver disease. Together, our findings establish that inherited disease risk is organized into latent regulatory programs, revealing a fundamental layer of biological heterogeneity underlying complex disease.

genomics↗

bioRxiv2026

Vgll2 and Tead1 Govern Generation of Mouse and Human Hypothalamic Hypocretin (Orexin) Neurons

The hypothalamic hypocretin (Hcrt; aka orexin) neuropeptide neurons are crucial for wakefulness and their malfunction results in narcolepsy; a devastating disorder without current cures. Transplantation of mouse Hcrt cells into mouse narcolepsy models is pointing to cell therapy as a potential treatment for narcolepsy. However, progress in this area requires decoding the developmental pathways generating human HCRT neurons and a deeper understanding of their diversity and physiology. Here, we identify the hypothalamic progenitor domain generating mouse Hcrt neurons and propose a conserved mouse/human genetic cascade driving Hcrt specification. This cascade involves the Vgll2 and Tead1 transcription factor partners, which we find are crucial for mouse Hcrt neuron generation. Conversely, co-misexpression of VGLL2 and TEAD1 in human stem cell derived hypothalamic organoids can induce HCRT neuron differentiation. These findings provide an important element for the development of human induced HCRT neurons for physiological studies and for cell therapy in narcolepsy.

neuroscience↗

bioRxiv2026

Leveraging TNFR2 for antitumour immunity: T reg depletion and myeloid reprogramming versus T cell costimulation

Although checkpoint inhibitors have revolutionized cancer treatment, responses are largely restricted to targeting CTLA-4 and PD-(L)1. TNFR2 has been identified as a target of interest, but a lack of understanding of whether agonists or blockers should be used, and whether Fc{gamma}R interactions promote effcacy, has hampered therapeutic antibody development. Here, we engineered two distinct -TNFR2 types: ligand-blocking Fc{gamma}R-engaging versus non-ligand-blocking agonist antibodies. Both showed potent anti-tumor effcacy across multiple syngeneic mouse tumor models and combined effectively with -PD-1. The ligand-blocking antibody depleted intratumoral T regs and reprogrammed the myeloid compartment, directing monocytes towards a pro-inflammatory macrophage and dendritic cell trajectory. Surprisingly, this effect depended on both inhibitory and activating Fc{gamma}Rs. In contrast, the agonist directly co-stimulated T and NK cells, through partially Fc{gamma}R-independent mechanisms. Both antibodies converged on activating tumor-specific CD8+ T cells mediating tumor rejection. Clinical assessment of both ligand-blocking (BI-1808) and agonist (BI-1910) -TNFR2 antibodies is currently ongoing. Graphical abstractO_LILigand-blocking anti-TNFR2 mAbs (BI-1808 and 3F10) deplete CCR8+ T regs and activate myeloid cells, remodeling the TME and leading to CD8+ T cell killing of cancer cells. T reg depletion and myeloid activation are mediated through interaction with activating Fc{gamma}Rs and the inhibitory Fc{gamma}RIIB, respectively. C_LIO_LIAgonist anti-TNFR2 mAbs (BI-1910 and 5A05) intrinsically agonize TNFR2 on T cells and NK cells, which is enhanced by Fc{gamma}RIIB crosslinking. The resulting broad lymphocyte activation mediates cancer cell killing. C_LI O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=127 SRC="FIGDIR/small/697572v1_ufig1.gif" ALT="Figure 1"> View larger version (30K): org.highwire.dtl.DTLVardef@3ecc17org.highwire.dtl.DTLVardef@c936a7org.highwire.dtl.DTLVardef@23189aorg.highwire.dtl.DTLVardef@243a05_HPS_FORMAT_FIGEXP M_FIG C_FIG

immunology↗

bioRxiv2025

Generating Hybrid Proteins with the MSA-Transformer

Protein superfamilies display extensive sequence and functional divergence, providing a rich landscape for engineering functionally enhanced variants. We present a stochastic, iterative framework that leverages the MSA-Transformer to generate intermediate sequences between a homologous pair of user-specified "source" and "target" proteins in sequence space. Sequence sites to "mask" is either selected by embedding-based dissimilarity or by row-attention information, while beam search concurrently explores multiple mutational pathways. Pre-trained sparse autoencoders combined with sequence and structural analyses are used to trace the inheritance and exchange of features across the "mutational pathway", revealing "hybrid" sequences that integrate properties of both source and target proteins. Applied across diverse protein families, the framework produces sequences that occupy biologically meaningful regions of sequence space and achieve higher consistency and plausibility scores than random baselines according to in silico metrics. In the B1/B2 metallo-{beta}-lactamase family, hybrids largely retain their core fold recombining structural and active-site motifs from both subclasses, demonstrating the models capacity to preserve catalytic features while exploring novel structural permutations. Code availabilityThe implementation is available via GitHub at https://github.com/santule/protmixy.

bioinformatics↗

bioRxiv2025

Squidly: Enzyme Catalytic Residue Prediction Harnessing a Biology-Informed Contrastive Learning Framework

Enzymes present a sustainable alternative to traditional chemical industries, drug synthesis, and bioremediation applications. Because catalytic residues are the key amino acids that drive enzyme function, their accurate prediction facilitates enzyme function prediction. Sequence similarity-based approaches such as BLAST are fast but require previously annotated homologs. Machine learning approaches aim to overcome this limitation; however, current gold-standard machine learning (ML)-based methods require high-quality 3D structures limiting their application to large datasets. To address these challenges, we developed Squidly, a sequence-only tool that leverages contrastive representation learning with a biology-informed, rationally designed pairing scheme to distinguish catalytic from non-catalytic residues using per-token Protein Language Model embeddings. Squidly surpasses state-of-the-art ML annotation methods in catalytic residue prediction while remaining sufficiently fast to enable wide-scale screening of databases. We ensemble Squidly with BLAST to provide an efficient tool that annotates catalytic residues with high precision and recall for both in- and out-of-distribution sequences.

bioinformatics↗

bioRxiv2025

Comprehensive molecular impact mapping of common and rare variants at GWAS loci

Deep learning sequence to function models can predict the molecular effects of genetic variants, but their predictions are limited to the cell types and assays they are trained on. Here we describe DNACipher, a deep learning model that predicts the effects of genetic variants across diverse biological contexts--including those not directly measured. DNACipher takes 196 kb of genome sequences as input and imputes variant effects across 38,582 cell type-assay combinations. DNACipher generates predictions for >7 times as many contexts as Enformer, which allows for better detection of variant effects at expression quantitative trait loci (eQTLs). We also introduce DNACipher Deep Variant Impact Mapping (DVIM), a method to identify variants with molecular effects at genome-wide association study (GWAS) loci. Application of DVIM to type 1 diabetes (T1D) reduced the mean fine-mapping credible set size from 24 to 1.4 variants per signal. DVIM variants had significantly higher fine-mapping posterior probabilities, and their predicted effects were supported by single-nucleus ATAC-seq and luciferase assays. DVIM also detected 6547 rare variants with molecular effects at 96% of GWAS T1D loci, and these were enriched for associations with immune traits. In summary, DNACipher DVIM prioritises common and rare variants at GWAS loci by predicting molecular effects across a broad range of contexts.

genetics↗

bioRxiv2025

Multilingual model improves zero-shot predictionof disease effects on proteins

Predicting variant effects remains a central challenge in genomics. Protein language models (PLMs) capture amino-acid-level sequence constraints, whereas codon language models operate on coding sequences and may retain information that is lost upon translation. Here, we tested whether scores from the codon language model CaLM provide predictive information beyond protein-level representations for missense-variant interpretation. Across 71,436 ClinVar missense variants from 11,554 genes, adding CaLM to PLM baselines produced modest but reproducible improvements under gene-held-out cross-validation. PLM-only ensemble controls and explicit mutational-context analyses indicated that this improvement could not be explained solely by generic ensembling or simple codon-substitution features. Aggregating CaLM probabilities across synonymous codons attenuated codon-degeneracy-associated discordance while preserving most of the broader differences between CaLM and PLM scores. Gene-level analyses further showed that CaLM contribution varied continuously across genes and depended partly on the protein-model background. Across ClinMAVE functional assays, however, improvements were less consistent, indicating that codon-protein complementarity is context dependent rather than universal. Together, these results identify a modest but reproducible component of variant-effect information in CaLM-derived codon-level scores that is not fully captured by protein-level language-model representations.

bioinformatics↗

bioRxiv2024

Learning the Language of Phylogeny with MSA Transformer

Classical phylogenetics assumes site independence, potentially overlooking epistasis. Protein language models capture dependencies in conserved structural and functional domains across the protein universe. Here, we ask whether MSA Transformer, which takes a multiple sequence alignment (MSA) as input, captures evolutionary distance and to what extent its representations reflect epistasis in protein sequence evolution, neither of which are explicitly available during training. Systematic shuffling of natural and simulated MSAs demonstrates the model exploits column-wise conservation to distinguish phylogenetic relationships. Using internal embeddings, we reconstruct trees that are markedly consistent with trees generated by maximum likelihood inference. Applying this approach to both the RNA-dependent RNA polymerase of RNA viruses and the nucleo-cytoplasmic large DNA virus domain, we recover both established and novel evolutionary relationships. We conclude that MSA Transformer complements, rather than replaces, classical inference for more accurate histories of protein families.

bioinformatics↗

bioRxiv2024

Epigenetic constraint of cellular genomes evolutionarily links genetic variation to function

Cellular diversity is a product of evolution acting to drive divergent regulatory programs from a common genome. Here, we use cross-cell-type epigenetic conservation to gain insight into the impact of selective constraints on genome function and phenotypic variation. By comparing chromatin accessibility across hundreds of diverse cell-types, we identify 1.4% of the human genome safeguarded by conserved domains of facultative heterochromatin, which we term regions under "cellular constraint". We calculate single-base resolution cellular constraint scores and demonstrate robust prediction of functionally important coding and non-coding loci in a cell-type-, trait-, and disease-agnostic manner. Cellular constraint annotation enhances causal variant identification, drug discovery, and clinical diagnostic predictions. Furthermore, cell-constrained sequences share paradoxical evolutionary signals of positive and negative selection, suggesting a dynamic role in driving human adaptation. Overall, this study demonstrates that evolutionary chromatin dynamics can be leveraged to inform the translation of genetic discoveries into effective biological, therapeutic, and clinical outcomes.

genetics↗

bioRxiv2024

Do Protein Language Models Learn Phylogeny?

Deep machine learning demonstrates a capacity to uncover evolutionary relationships directly from protein sequences, in effect internalising notions inherent to classical phylogenetic tree inference. We connect these two paradigms by assessing the capacity of protein-based language models (pLMs) to discern phylogenetic relationships without being explicitly trained to do so. We evaluate ESM2, ProtTrans and MSA-Transformer relative to classical phylogenetic methods, while also considering sequence insertions and deletions (indels) across 114 Pfam datasets. The largest ESM2 model tends to outperform other pLMs (including the multimodal ESM3) by recovering phylogenetic relationships among homologous protein sequences in both low- and high-gap settings. pLMs agree with conventional phylogenetic methods in general, but more so for protein families with fewer implied indels, highlighting indels as a key factor differentiating classical phylogenetics from pLMs. We find that pLMs preferentially capture broader as opposed to finer evolutionary relationships within a specific protein family, where ESM2 has a sweet spot for highly divergent sequences, at remote distance. Less than 10% of neurons are sufficient to broadly recapitulate classical phylogenetic distances; when used in isolation the difference between the paradigms is further diminished. We show these neurons are polysemantic, shared among different homologous families but never fully overlapping. We highlight the potential of ESM2 as a complementary tool for phylogenetic analysis, especially when extending to remote homologs that are difficult to align and imply complex histories of insertions and deletions.

molecular biology↗

bioRxiv2024

Optimal Phylogenetic Reconstruction of Insertion and Deletion Events

Insertions and deletions (indels) influence the genetic code in fundamentally distinct ways from substitutions, significantly impacting gene product structure and function. Despite their influence, the evolutionary history of indels is often neglected in phylogenetic tree inference and ancestral sequence reconstruction, hindering efforts to comprehend biological diversity determinants and engineer variants for medical and industrial applications. We frame determining the optimal history of indel events as a single Mixed-Integer Programming (MIP) problem, across all nodes in a phylogenetic tree adhering to topological constraints, and all sites implied by a given set of aligned, extant sequences. By disentangling the impact on ancestral sequences at each branch point, this approach identifies the minimal indel events that jointly explain the diversity in sequences mapped to the tips of that tree. MIP can recover alternate optimal indel histories, if available. We evaluated MIP for indel inference on a dataset comprising 15 real phylogenetic trees associated with protein families ranging from 165 to 2000 extant sequences, and on 60 synthetic trees at comparable scales of data and reflecting realistic rates of mutation. Across relevant metrics, MIP outperformed alternative parsimony-based approaches and reported the fewest indel events, on par or below their occurrence in synthetic datasets. MIP offers a rational justification for indel patterns in extant sequences; importantly, it uniquely identifies global optima on complex protein data sets without making unrealistic assumptions of independence or evolutionary underpinnings, promising a deeper understanding of molecular evolution and aiding novel protein design.

bioinformatics↗

bioRxiv2022

Inferring cell diversity in single cell data using consortium-scale epigenetic data as a biological anchor for cell identity

Methods for cell clustering and gene expression from single-cell RNA sequencing (scRNA-seq) data are essential for biological interpretation of cell processes. Here we present TRIAGE-Cluster which uses genome-wide epigenetic data from diverse bio-samples to identify genes demarcating cell diversity in scRNA-seq data. TRIAGE-Cluster integrates patterns of repressive chromatin deposited across diverse cell types with weighted density estimation to determine cell type clusters in a 2D UMAP space. We then present TRIAGE-ParseR, a machine learning method that evaluates gene expression rank lists to define gene groups governing the identity and function of cell types. We demonstrate the utility of this two-step approach using atlases of in vivo and in vitro cell diversification and organogenesis. We also provide a web accessible dashboard for analysis and download of data and software. Collectively, genome-wide epigenetic repression provides a versatile strategy to define cell diversity and study gene regulation of scRNA-seq data.

bioinformatics↗

bioRxiv2022

An integrated cell barcoding and computational analysis pipeline for scalable analysis of differentiation at single-cell resolution

Pluripotent stem cells provide a scalable approach to analyse molecular regulation of cell differentiation across multiple developmental lineage trajectories. In this study, we engineered barcoded iPSCs to generate an atlas of multilineage differentiation from pluripotency, encompassing a time-course of WNT-induced differentiation perturbed using modulators of WNT, BMP, and VEGF signalling. Computational mapping of in vitro cell types to in vivo developmental lineages revealed a diversity of iPSC-derived cell types comprising mesendoderm lineage cell types including lateral plate and paraxial mesoderm, neural crest, and primitive gut. Coupling this atlas of in vitro differentiation with Summary data-based Mendelian Randomisation analysis of human complex traits, we identify the WNT-inhibitor protein TMEM88 as a putative regulator of mesendodermal cell types governing development of diverse cardiovascular and anthropometric traits. Using genetic loss of function models, we show that TMEM88 is required for differentiation of diverse endoderm and mesoderm cell lineages in vitro and that TMEM88 knockout in vivo results in a significant dysregulation of arterial blood pressure. This study provides an atlas of multilineage iPSC differentiation coupled with new molecular, computational, and statistical genetic tools to dissect genetic determinants of mammalian developmental physiology.

genomics↗

bioRxiv2022

Cytocipher detects significantly different populations of cells in single cell RNA-seq data

Identification of cell types using single cell RNA-seq (scRNA-seq) is revolutionising the study of multicellular organisms. However, typical scRNA-seq analysis often involves post hoc manual curation to ensure clusters are transcriptionally distinct, which is time-consuming, error-prone, and irreproducible. To overcome these obstacles, we developed Cytocipher, a bioinformatics method and scverse compatible software package that statistically determines significant clusters. Application of Cytocipher to normal tissue, development, disease, and large-scale atlas data reveals the broad applicability and power of Cytocipher to generate biological insights in numerous contexts. This included the identification of cell types not previously described in the datasets analyzed, such as CD8+ T cell subtypes in human peripheral blood mononuclear cells; cell lineage intermediate states during mouse pancreas development; and subpopulations of luminal epithelial cells over-represented in prostate cancer. Cytocipher also scales to large datasets with high test performance, as shown by application to the Tabula Sapiens Atlas representing >480,000 cells. Cytocipher is a novel and generalisable method that statistically determines transcriptionally distinct and programmatically reproducible clusters from single cell data. Cytocipher is available at https://github.com/BradBalderson/Cytocipher.

bioinformatics↗

bioRxiv2022

SiRCle (Signature Regulatory Clustering) model integration reveals mechanisms of phenotype regulation in renal cancer

Clear cell renal cell carcinoma (ccRCC) tumours develop and progress via complex remodelling of the kidney epigenome, transcriptome, proteome, and metabolome. Given the subsequent tumour and inter-patient heterogeneity, drug-based treatments report limited success, calling for multi-omics studies to extract regulatory relationships, and ultimately, to develop targeted therapies. However, current methods are unable to extract nonlinear multi-omics perturbations. Here, we present SiRCle (Signature Regulatory Clustering), a novel method to integrate DNA methylation, RNA-seq and proteomics data. Applying SiRCle to a case study of ccRCC, we disentangle the layer (DNA methylation, transcription and/or translation) where dys-regulation first occurs and find the primary biological processes altered. Next, we detect regulatory differences between patient subsets by using a variational autoencoder to integrate omics data followed by statistical comparisons on the integrated space. In ccRCC patients, SiRCle allows to identify metabolic enzymes and cell-type-specific markers associated with survival along with the likely molecular driver behind the genes perturbations.

cancer biology↗