Alzheimer's disease (AD) is a complex, multifactorial neurodegenerative disorder with substantial heterogeneity in progression and treatment response. Despite recent therapeutic advances, predictive models capable of accurately forecasting individualized future biomarker states remain limited. Here, we present a machine learning-based operator learning framework for personalized modeling of AD progression, integrating longitudinal multimodal imaging, biomarker, and clinical data. Unlike conventional models with prespecified dynamics, our approach directly learns patient-specific disease operators governing the spatiotemporal evolution of amyloid, tau, and neurodegeneration biomarkers. Using Laplacian eigenfunction bases, we construct geometry-aware neural operators capable of capturing complex brain dynamics. Embedded within a digital twin paradigm, the framework enables individualized predictions, simulation of therapeutic interventions, and in silico clinical trials. Applied to AD clinical data, our method achieves high prediction accuracy exceeding 90% across multiple biomarkers, substantially outperforming existing approaches. This work offers a scalable, interpretable platform for precision modeling and personalized therapeutic optimization in neurodegenerative diseases.

Epithelial-to-Mesenchymal Transition (EMT) is a critical biological process by which cells acquire enhanced migratory and invasive properties. A key signaling pathway involved in EMT phenotypes includes transforming growth factor β (TGFβ) and transcription factors (TFs) such as SNAIL, ZEB, and TWIST. Additionally, microRNAs (miRNAs) - small, non-coding molecules that regulate gene expression by targeting mRNA transcripts - directly regulate genes central to the EMT process. Notably, miR-22 has been identified as a significant regulator of EMT through direct inhibition of EMT drivers like SNAI1 and indirect regulation of upstream genes. In this study, we performed CRISPR-based network rewiring by selectively removing an edge-the connection between two nodes-to investigate its impact on EMT dynamics. Specifically, we disrupted the connection between miR-22 and SNAI1 without affecting other interactions involving miR-22 or SNAI1 and examined the resulting effects on EMT. We demonstrate that the removal of the miR-22 target site from the SNAI1 gene renders cells more sensitive to TGFβ-mediated EMT. This finding highlights the unique advantage of edge-specific perturbation by ablating the direct regulatory connection between miR-22 and SNAI1. We demonstrate that all measured downstream effects on EMT can be attributed to this single interaction, independent of miR-22's influence on other targets or indirect pathways. More generally, our results underscore the importance of CRISPR-mediated edge ablation for exploring the interactions that govern biological networks and highlight an underexplored opportunity to develop edge-based therapeutic modalities.

Saccharomyces cerevisiae is a widely used biotechnological workhorse in both academic and industrial settings. One reason for its continued popularity is the extensive legacy of genetic tools, developed over its long history of use, that enable precise manipulation of the S. cerevisiae genome. These tools have enabled extensive genetic characterization and dramatic re-programming efforts for applications ranging from fundamental research to industrial chemical production. However, existing software is often designed for automated pipelines and can be cumbersome to integrate into existing laboratories. Here we present a digital toolkit called PYEAST (Python Enabled Automated Strain Transformation) that encodes some of the most widely used methods for working with S. cerevisiae, modernizes them to leverage advances in DNA synthesis, and facilitates digital sequence management.

Major neuropsychiatric disorders such as major depressive disorder (MDD) and schizophrenia (SCZ), as well as the neurodevelopmental disorder autism spectrum disorder (ASD), are traditionally treated as distinct clinical entities. However, genome-wide association studies indicate shared genetic risks, motivating a transdiagnostic view. Resting-state functional connectivity (rsFC) is a promising biomarker for these disorders, but its high dimensionality complicates inference of inter-disorder relationships in the native feature space. Here, we develop an rsFC-based embedding-relation workflow that quantifies disorder relationships in a connectivity-informed, low-dimensional embedding space. Central to the workflow is a mutual information-based embedding framework that evaluates candidate embedding approaches and selects an optimal strategy. Using synthetic connectivity data, the framework indicates that rsFC embeddings are best represented in a spherical space under a moderate level of supervision. Building on this insight, we applied the workflow to curated, multi-disorder rsFC datasets to derive shared embedding spaces encompassing the connectivity features of ASD, MDD, and SCZ. In these spaces, we consistently observed a robust three-way relationship: a pronounced neurobiological dissimilarity between ASD and MDD, contrasted with greater similarity between SCZ and both disorders. These findings support a dimensional, transdiagnostic perspective on neuropsychiatric disorders and offer new insights into their shared and distinct neural underpinnings.

Metabolic network models are widely used in systems and synthetic biology to study cellular metabolism from microbes to mammals. However, their application to mammalian cells remains limited by the scarcity of intracellular flux measurements relative to network size. As a result, mammalian metabolic models are typically underdetermined and require optimisation-based assumptions to be solved. Unlike microbial systems, mammalian cells do not generally operate under growth-maximising objectives, making objective specification particularly challenging. To address this limitation, we propose CellTarget, an algorithm for inferring cellular objectives directly from experimental flux data. CellTarget combines a convex Flux Balance Analysis (FBA) module with an error-minimisation module that iteratively adjusts objective coefficients to improve agreement between predicted and measured fluxes. Gradients are obtained via implicit differentiation of the convex FBA optimisation problem, enabling backpropagation of the inferred objective coefficients and joint optimisation of both modules. CellTarget is evaluated across producer and non-producer Chinese Hamster Ovary (CHO) cell lines, growth phases, objective dimensionalities and both reduced and genome-scale networks. Across all settings, CellTarget yields accurate forward flux predictions and outperforms a benchmark inverse FBA method. The results reveal substantial non-identifiability of cellular objectives, with predictive accuracy driven more strongly by network constraints than by objective specificity.

Myofibril arrangement is critical to cardiac muscle function in health and disease. Historically, analysis of the impact of myofibril organisation on force and cell contraction has relied on the assumption of uniaxial arrays. However, improvements in imaging indicate that myofibrils form complex networks, though how these networks modulate force has yet to be explored. Here, morphological analysis of sheep left-ventricular cardiomyocytes is utilised to inform a non-linear finite element model of cell contraction. Analysis of deep learning segmentations of z-discs demonstrate that myofibrils are oriented about the contraction axis (mean ) but deviate locally by up to (standard deviation ). Simulations produce unique deformations for geometries informed by myofibril orientations, displaying internal rotation and off-axis deformations. Moreover, anisotropy generates shear stresses distinct from the uniaxial case, demonstrating spatial relationships that balance shear across the cell and a correlation between shear stress and z-disc orientation. These findings highlight the impact of myofibril networks on forces during cell contraction.

In biotechnological processes, cell density and physiology are critical parameters for controlling the feed rate, harvest time, and process performance. We developed an automated flow cytometry approach that enables continuous, real-time (fully automated, hourly) monitoring of bacterial populations in continuous bioreactors. The method employed a double-staining protocol that combined DAPI to assess total DNA content and Alexa Fluor 488-EdU via Click-iT technology to identify the proportions of cells undergoing active DNA replication through EdU incorporation. The integrated workflow included fixation, permeabilization, staining, and measurement steps and was applied to three Gram-negative strains: Bradyrhizobium sp., Escherichia coli, and Stenotrophomonas rhizophila. Automated analysis captured growth dynamics and cell cycle progression, providing insights into population behavior under different dilution rates. In this study, automated on-line sampling enabled hourly flow cytometry measurements of cell concentration and physiological indicators during continuous cultivation, supporting real-time monitoring and control in industrial biotechnology.

Head and neck squamous cell carcinoma (HNSCC) is a complex multivariable disease posing a significant challenge in therapeutics. OXPHOS and PPP are centrally implicated in metabolic heterogeneity influencing tumor behavior, treatment response, and patient outcomes. This study aims to stratify HNSCC based on OXPHOS and PPP gene expression to construct a clinical outcome risk predictive model. HNSCC patients were stratified into four metabolic subtypes including mixed, OXPHOS-leaning, PPP-leaning, and quiescent. The quiescent metabolic subtype showed longest survival with lower proliferation scores, whereas mixed subtype showed the worst survival with higher proliferation scores. Metabolic proteins of the CPTAC and ICPC cohorts affirmed the existence of metabolic heterogeneity among tumor samples. A prognostic risk predictive model was developed based on thirteen genes, which performed better than the OXPHOS-PPP-glycolysis model and other predictive forty-seven metabolic gene’s model. Existence of metabolic heterogeneity was successfully determined in HNSCC. OXPHOS and PPP genes enriched in mixed metabolic subtype having the worst clinical outcome are suggestive of higher metastatic potential, which might offer advisement in personalized therapeutics.

Living Biotherapeutic Products (LBPs) offer a promising therapeutic strategy for metabolic disorders rooted in gut microbiome dysfunction, yet quantitative frameworks for predicting their efficacy remain underdeveloped. We introduce the Bacterial Compartment Absorption and Transit (BCAT) model, a pharmacokinetic-pharmacodynamic framework that couples probiotic transit, endogenous microbiome metabolism, and enzymatic transformation within a unified dose-optimization setting. Building on the classical CAT model, BCAT incorporates mechanistically-derived colon compartments and treats dosing time as a control variable. We validate BCAT against clinical data for native choline metabolism and SYNB1618 probiotic trials, achieving 5% mean prediction error compared to ~30% for prior two-compartment models. Applying BCAT to trimethylaminuria (TMAU), we predict that ~109 CFU of engineered probiotic, administered 3-4 h before meals, achieves 95% reduction in systemic trimethylamine, matching healthy hepatic clearance. Global sensitivity analysis identifies enzyme expression level as the dominant design parameter, enforcing the broad applicability of this model. The BCAT framework generalizes to any gut microbiome-mediated metabolic disorder and provides quantitative dosing targets to guide live biotherapeutic development.

Acute myeloid leukemia (AML) is a hematologic malignancy originating in the bone marrow and often progressing to extramedullary sites. Despite advances in molecularly targeted therapies and hematopoietic stem cell transplantation, clinical outcomes remain poor. Tyrosine kinase inhibitors (TKIs) provide benefit to a subset of AML patients harboring FLT3-ITD mutations; however, relapse and resistance remain common. These therapeutic failures are driven by both intrinsic properties of leukemic stem cells (LSCs)—a quiescent, self-renewing population—and extrinsic cues from the tumor microenvironment. We previously demonstrated that arteriolar endothelial cells (ECs) produce miR-126, which is transferred to LSCs, promoting quiescence, treatment resistance, and niche retention. During disease progression, TNF-α secreted by expanding blasts suppresses EC miR-126 production. Following TKI administration, blast reduction lowers TNF-ɑ levels, restoring EC miR-126 production, and this miR-126 expression enables LSCs to re-enter quiescence—thereby escaping therapy and facilitating relapse. To explore this dynamic, we developed an agent-based computational model of the AML bone marrow microenvironment, parameterized with in vitro and in vivo data. The model captures vascular niche remodeling and feedback between leukemic populations and endothelial signaling. Simulations reveal that LSC protection mediated by miR-126 can be disrupted by combining TKIs with miRisten, a miR-126 inhibitor. When administered on a defined schedule, this combination dismantles the protective niche and enhances LSC eradication. These findings underscore the therapeutic potential of targeting microenvironmental feedback to overcome resistance and prevent AML relapse.