Explore the vast range of titles available.

Compounds with defined multi-target activity play an increasingly important role in drug discovery. Structural features that might be signatures of such compounds have mostly remained elusive thus far. We have explored the potential of explainable machine learning to uncover structural motifs that are characteristic of dual-target compounds. For a pharmacologically relevant target pair-based test system designed for our study, accurate prediction models were derived and the influence of molecular representation features of test compounds was quantified to explain the predictions. The analysis revealed small numbers of specific features whose presence in dual-target and absence in single-target compounds determined accurate predictions. These features formed coherent substructures in dual-target compounds. From computational analysis of specific feature contributions, structural motifs emerged that were confirmed to be signatures of different dual-target activities. Our findings demonstrate the ability of explainable machine learning to bridge between predictions and intuitive chemical analysis and reveal characteristic substructures of dual-target compounds.

Universal Differential Equations (UDEs) combine mechanistic differential equations with data-driven artificial neural networks, forming a flexible framework for modelling complex biological systems. This hybrid approach leverages prior knowledge and data to uncover unknown processes and deliver accurate predictions. However, UDEs face challenges in efficient and reliable training due to stiff dynamics and noisy, sparse data common in biology, and in ensuring the interpretability of the parameters of the mechanistic model. We investigate these challenges and evaluate UDE performance on realistic biological scenarios, providing a systematic training pipeline. Our results demonstrate the versatility of UDEs in systems biology and reveal that noise and limited data significantly degrade performance, but regularisation can improve accuracy and interpretability. By addressing key challenges and offering practical solutions, this work advances UDE methodology and underscores its potential in tackling complex problems in systems biology.

Oats have various positive effects on human health, but the underlying mechanisms are not fully understood. To identify oat-microbiome-host interactions contributing to metabolic improvements, we conducted two randomized controlled dietary interventions in parallel-design in individuals with metabolic syndrome, comparing a short-term, high-dose and a six-week, moderate oat intake with respective controls (DRKS00022169). Both oat diets lead to an increase in plasma ferulic acid (0.64 [0.26, 1.02], P  = 0.002; 0.55 [0.21, 0.89], P  = 0.003), while the high-dose oat-diet also increased dihydroferulic acid (1.23 [0.44, 2.01], P  = 0.003). Here we show that microbial phenolic metabolites are driving factors for the cholesterol-lowering effect of oats, which might be of relevance since short-term, high-dose oat-diet is a suitable approach to alleviate obesity-related lipid disorders. Oat consumption, particularly a short-term, high-dose oat diet, improves metabolic health in individuals with metabolic syndrome by increasing microbially produced phenolic metabolites, leading to lower blood cholesterol levels.

Universal differential equations (UDEs) leverage the respective advantages of mechanistic models and artificial neural networks and combine them into one dynamic model. However, these hybrid models can suffer from unrealistic solutions, such as negative values for biochemical quantities. We present non-negative UDE (nUDEs), a constrained UDE variant that guarantees non-negative values. Furthermore, we explore regularisation techniques to improve generalisation and interpretability of UDEs.

Universal Differential Equations (UDEs) combine mechanistic differential equations with data-driven artificial neural networks, forming a flexible framework for modelling complex biological systems. This hybrid approach leverages prior knowledge and data to uncover unknown processes and deliver accurate predictions. However, UDEs face challenges in efficient and reliable training due to stiff dynamics and noisy, sparse data common in biology, and in ensuring the interpretability of the parameters of the mechanistic model. We investigate these challenges and evaluate UDE performance on realistic biological scenarios, providing a systematic training pipeline. Our results demonstrate the versatility of UDEs in systems biology and reveal that noise and limited data significantly degrade performance, but regularisation can improve accuracy and interpretability. By addressing key challenges and offering practical solutions, this work advances UDE methodology and underscores its potential in tackling complex problems in systems biology.Competing Interest StatementThe authors have declared no competing interest.

Summary Dendrites display a striking variety of neuronal type-specific morphologies, but the mechanisms and principles underlying such diversity remain elusive. A major player in defining the morphology of dendrites is the neuronal cytoskeleton, including evolutionarily conserved actin-modulatory proteins (AMPs). Still, we lack a clear understanding of how AMPs might support developmental phenomena such as neuron-type specific dendrite dynamics. To address precisely this level of in vivo specificity, we concentrated on a defined neuronal type, the class III dendritic arborisation (c3da) neuron of Drosophila larvae, displaying actin-enriched short terminal branchlets (STBs). Computational modelling reveals that the main branches of c3da neurons follow a general growth model based on optimal wiring, but the STBs do not. Instead, model STBs are defined by a short reach and a high affinity to grow towards the main branches. We thus concentrated on c3da STBs and developed new methods to quantitatively describe dendrite morphology and dynamics based on in vivo time-lapse imaging of mutants lacking individual AMPs. In this way, we extrapolated the role of these AMPs in defining STB properties. We propose that dendrite diversity is supported by the combination of a common step, refined by a neuron type-specific second level. For c3da neurons, we present a molecular model of how the combined action of multiple AMPs in vivo define the properties of these second level specialisations, the STBs. In brief A quantitative morphological dissection of the concerted actin-modulatory protein actions provides a model of dendrite branchlet outgrowth. Highlights * Actin organisation in small terminal branchlets of Drosophila class III dendritic arborisation neurons * Six actin-modulatory proteins individually control the characteristic morphology and dynamics of branchlets * Quantitative tools for dendrite morphology and branch dynamics enable a comparative analysis * A two-step computational growth model reproduces c3da dendrite morphology Competing Interest Statement The authors have declared no competing interest. Footnotes * ↵1 Senior authors * ↵2 Lead contact