Phil Ostheimer, Mayank Nagda, Andriy Balinskyy
et al.
Sparse data is fundamental to scientific simulations in biology and physics, from single-cell gene expression to particle calorimetry, where exact zeros encode physical absence rather than weak signal. However, existing diffusion models lack the physical rigor to faithfully represent this sparsity. This work introduces Sparse Data Diffusion (SDD), a generative method that explicitly models exact zeros via Sparsity Bits, unifying efficient ML generation with physically grounded sparsity handling. Empirical validation in particle physics and single-cell biology demonstrates that SDD achieves higher fidelity than baseline methods in capturing sparse patterns critical for scientific analysis, advancing scalable and physically faithful simulation.
Background: Large foundation models have revolutionized single-cell analysis, yet no kidney-specific model currently exists, and it remains unclear whether organ-focused models can outperform generalized models. The kidney's complex cellular architecture further complicate integration of large-scale omics data, where current frameworks trained on limited datasets struggle to correct batch effects, capture cross-modality variation, and generalize across species. Methods: We developed Nephrobase Cell+, the first kidney-focused large foundation model, pretrained on ~100 billion tokens from ~39.5 million single-cell and single-nucleus profiles across 4,319 samples. Nephrobase Cell+ uses a transformer-based encoder-decoder architecture with gene-token cross-attention and a mixture-of-experts module for scalable representation learning. Results: Nephrobase Cell+ sets a new benchmark for kidney single-cell analysis. It produces tightly clustered, biologically coherent embeddings in human and mouse kidneys, far surpassing previous foundation models such as Geneformer, scGPT, and UCE, as well as traditional methods such as PCA and autoencoders. It achieves the highest cluster concordance and batch-mixing scores, effectively removing donor/assay batch effects while preserving cell-type structure. Cross-species evaluation shows superior alignment of homologous cell types and >90% zero-shot annotation accuracy for major kidney lineages in both human and mouse. Even its 1B-parameter and 500M variants consistently outperform all existing models. Conclusions: Nephrobase Cell+ delivers a unified, high-fidelity representation of kidney biology that is robust, cross-species transferable, and unmatched by current single-cell foundation models, offering a powerful resource for kidney genomics and disease research.
Izzy Newsham, Luka Kovačević, Richard Moulange
et al.
Genes, proteins and other biological entities influence one another via causal molecular networks. Causal relationships in such networks are mediated by complex and diverse mechanisms, through latent variables, and are often specific to cellular context. It remains challenging to characterise such networks in practice. Here, we present a novel framework to evaluate large language models (LLMs) for zero-shot inference of causal relationships in biology. In particular, we systematically evaluate causal claims obtained from an LLM using real-world interventional data. This is done over one hundred variables and thousands of causal hypotheses. Furthermore, we consider several prompting and retrieval-augmentation strategies, including large, and potentially conflicting, collections of scientific articles. Our results show that with tailored augmentation and prompting, even relatively small LLMs can capture meaningful aspects of causal structure in biological systems. This supports the notion that LLMs could act as orchestration tools in biological discovery, by helping to distil current knowledge in ways amenable to downstream analysis. Our approach to assessing LLMs with respect to experimental data is relevant for a broad range of problems at the intersection of causal learning, LLMs and scientific discovery.
Data-driven discovery of model equations is a powerful approach for understanding the behavior of dynamical systems in many scientific fields. In particular, the ability to learn mathematical models from data would benefit systems biology, where the complex nature of these systems often makes a bottom up approach to modeling unfeasible. In recent years, sparse estimation techniques have gained prominence in system identification, primarily using parametric paradigms to efficiently capture system dynamics with minimal model complexity. In particular, the Sindy algorithm has successfully used sparsity to estimate nonlinear systems by extracting from a library of functions only a few key terms needed to capture the dynamics of these systems. However, parametric models often fall short in accurately representing certain nonlinearities inherent in complex systems. To address this limitation, we introduce a novel framework that integrates sparse parametric estimation with nonparametric techniques. It captures nonlinearities that Sindy cannot describe without requiring a priori information about their functional form. That is, without expanding the library of functions to include the one that is trying to be discovered. We illustrate our approach on several examples related to estimation of complex biological phenomena.
Glycerol dialkyl glycerol tetraethers (GDGTs) derived from microorganisms are important tools for the study of paleoclimate changes. Incubation experiments are helpful to clarify the mechanisms for the responses of GDGTs to environmental parameters, and to test the reliability of related climatic proxies. However, previous GDGT incubation experiments were mainly conducted on a single strain or suffered from the influence of a background signal, hampering systematically understanding the precise response of this biomarker to environmental factors in a soil environment. In this paper, artificial soils without GDGTs were incubated under the same temperature but different soil water content (SWC) conditions. The results showed that: (1) The abundances of GDGTs were positively correlated with SWC, but phosphate buffer could inhibit the production of GDGTs; (2) The branched and isoprenoid tetraether index (BIT), a soil moisture proxy developed in natural soils, was not significantly correlated with SWC; (3) 6-methyl brGDGTs were more abundant than 5-methyl brGDGTs, resulting in extremely high values of MBT'5ME and low MBT'. The results suggest that the BIT soil moisture proxy may indirectly (rather than directly) respond to SWC changes and confirm that high relative abundance of 6-methyl brGDGTs can affect the applicability of the MBT'5ME paleothermometer in soils. The BRIEF REPORT is available for this paper at http://www.ykcs.ac.cn/en/article/doi/10.15898/j.ykcs.202405240120.
Jonathan M. Chase, A. Jeliazkov, E. Ladouceur
et al.
Metacommunity ecology combines local (e.g., environmental filtering and biotic interactions) and regional (e.g., dispersal and heterogeneity) processes to understand patterns of species abundance, occurrence, composition, and diversity across scales of space and time. As such, it has a great potential to generalize and synthesize our understanding of many ecological problems. Here, we give an overview of how a metacommunity perspective can provide useful insights for conservation biology, which aims to understand and mitigate the effects of anthropogenic drivers that decrease population sizes, increase extinction probabilities, and threaten biodiversity. We review four general metacommunity processes—environmental filtering, biotic interactions, dispersal, and ecological drift—and discuss how key anthropogenic drivers (e.g., habitat loss and fragmentation, and nonnative species) can alter these processes. We next describe how the patterns of interest in metacommunities (abundance, occupancy, and diversity) map onto issues at the heart of conservation biology, and describe cases where conservation biology benefits by taking a scale‐explicit metacommunity perspective. We conclude with some ways forward for including metacommunity perspectives into ideas of ecosystem functioning and services, as well as approaches to habitat management, preservation, and restoration.
Kartik Majila, Shreyas Arvindekar, Muskaan Jindal
et al.
Integrative modeling enables structure determination for large macromolecular assemblies by combining data from multiple sources of experiment data with theoretical and computational predictions. Recent advancements in AI-based structure prediction and electron cryo-microscopy have sparked renewed enthusiasm for integrative modeling; structures from AI-based methods can be integrated with in situ maps to characterize large assemblies. This approach previously allowed us and others to determine the architectures of diverse macromolecular assemblies, such as nuclear pore complexes, chromatin remodelers, and cell-cell junctions. Experimental data spanning several scales was used in these studies, ranging from high-resolution data, such as X-ray crystallography and Alphafold structures, to low-resolution data, such as cryo-electron tomography maps and data from co-immunoprecipitation experiments. Two recurrent modeling challenges emerged across a range of studies. First, modeling disordered regions, which constituted a significant portion of these assemblies, necessitated the development of new methods. Second, methods needed to be developed to utilize the information from cryo-electron tomography, a timely challenge as structural biology is increasingly moving towards in situ characterization. Here, we recapitulate recent developments in the modeling of disordered proteins and the analysis of cryo-electron tomography data and highlight opportunities for method development in the context of integrative modeling.
Zooplankton is the important component of aquatic ecosystems. These organisms are important biological indicator of water quality of aquatic ecosystem due to their response to the environmental changes. In this study, we investigated distribution of zooplankton biomass in the Shatt Al-Basra Canal and Shatt Al-Arab River. Zooplankton samples were collected from two stations in the Shatt Al-Basra Canal, before (S1) and after (S2) the dam, and two stations in the Shatt Al-Arab River, Al-Siba (S3) and Al-Faw (S4). The biomass of zooplankton in the Shatt Al-Basra Canal varied between 23.102 - 520.875 mg/m3 in terms of wet weight and 3.787 - 102.132 mg/m3 in terms of dry weight at two stations (before the dam and after the dam) during the period of January and May, respectively. The displacement volume and standing crops also showed variations of the biomass of zooplankton. In the Shatt Al-Basra Canal, the range was from 0.06 ml/m3 and 3.9 mgC/m3 during January at S1 to 1.083 ml/m3 and 70.395 mgC/m3 during May at S2. While in the Shatt Al-Arab River, the biomass of zooplankton in terms of wet weight ranged from 10.671 - 655.78 mg/m3 during December at S3 (Al-Siba) and may at S4 (Al-Faw) respectively. In terms of dry weight, the biomass ranged from 1.423 to 168.149 mg/m3 in S3 during the December and in S4 during May respectively. In terms of displacement volume and standing crops, they ranged from 0.03 ml/m3 to 1.95 mgC/m3 during December at S3 to 1.819 ml/m3 and 118.235 mgC/m3 during February at S4.
Abstract Daphnia can avoid predation by sensing fish kairomones and producing inducible defenses by altering the phenotype. In this study, the results showed that the morphological and life‐history strategies of two Daphnia species (Daphnia pulex and Daphnia sinensis) exposed to Aristichthys nobilis kairomones. In the presence of fish kairomones, the two Daphnia species exhibited significantly smaller body length at maturity, smaller body length of offspring at the 10th instar, and longer relative tail spine of offspring. Nevertheless, other morphological and life‐history traits of the two Daphnia species differed. D. pulex showed a significantly longer relative tail spine length and earlier age at maturity after exposure to fish kairomones. The total offspring number of D. sinensis exposed to fish kairomones was significantly higher than that of the control group, whereas that of D. pulex was significantly lower. These results suggest that the two Daphnia species have different inducible defense strategies (e.g., morphological and life‐history traits) during prolonged exposure to A. nobilis kairomones, and their offspring also develop morphological defenses to avoid predation. It will provide reference for further exploring the adaptive evolution of Daphnia morphology and life‐history traits in the presence of planktivorous fish.
Abstract Hepatocellular carcinoma (HCC) is recognized as a highly malignant tumor. Targeted combination immunotherapy, the initially approved regimen, is compromised by adverse side effects and low response rates during clinical treatment. Traditional Chinese medicine and its derived natural compounds, known for their anticancer effects, offer advantages of low toxicity and cost. In this study, we performed high-throughput phenotypic screening in vitro to identify promising anti-HCC drugs. Among 1,444 bioactive compounds, digoxigenin (DIG) was found to significantly impede HCC cell progression. We validated DIG’s therapeutic effects through assays such as cell counting by CCK8, lactate dehydrogenase, and colony formation. Analyses including transmission electron microscopy, western blotting, and immunofluorescence demonstrated that DIG inhibits HCC cell proliferation via autophagy. Network pharmacology and molecular docking studies suggest that DIG targets the PI3K/AKT/mTOR signaling pathway. Comparative treatments of Hep3B and Huh7 cells with DIG or mTOR inhibitors revealed similar inhibitory impacts, indicating that DIG induces autophagy by inhibiting the PI3K/AKT/mTOR pathway. In vivo studies confirmed that DIG halts the growth of subcutaneous xenograft tumors. In conclusion, DIG represents a potential HCC treatment by modulating the PI3K/AKT/mTOR pathway to induce autophagy. This research, via phenotypic screening, accelerates drug discovery and the development of novel therapies targeting the underlying mechanisms of liver cancer.
Neoplasms. Tumors. Oncology. Including cancer and carcinogens, Cytology
Jing Lu, Alexandra Garzon-Garcia, Ann Chuang
et al.
Point- and non-point source nutrients are likely to have different ecological impacts in receiving waters, due to differences in the concentration and proportions of nutrient fractions. However, the direct comparison of their ecological impacts in receiving waters has barely been quantified. We undertook algal bioassays with algal communities from river sites and showed that there was a photosynthetic yield (Fv/Fm) response to nutrient enrichment when river nutrient concentrations were relatively low, but not at higher nutrient concentrations. To combat this variability in the photosynthetic state of algae, we developed a standardized algal bioassay (3-day), using a cultured species of algae which was starved of nitrogen, to compare the photosynthetic response to three nitrogen sources: treated wastewater, aquaculture farm discharges, and soil erosion-derived nutrient sources. This study showed that the nutrient parameter that had the highest correlation with algal photosynthetic response was total dissolved nitrogen (TDN), i.e., the sum of dissolved inorganic and organic nitrogen, rather than dissolved inorganic nitrogen alone. This was true across all three nutrient sources (R2 = 0.58–0.78). Additionally, the same concentrations of TDN from soil erosion-derived and aquaculture samples resulted in a significantly higher algal photosynthetic response, compared to the treated wastewater. This indicates that TDN from soils and aquaculture farms was significantly more bioavailable to the cultured algae than treated wastewater. When a range of parameters were correlated with algal responses, organic and inorganic nutrients, and organic carbon were the parameters that had the highest explanatory power for soil erosion-derived and aquaculture samples (R2 = 0.75–0.87). The importance of organic compounds in these equations points to the potential of microbial transformation of organic nutrients into more bioavailable forms during the 3-day bioassay. This highlights the need to understand the relationship between algal and microbial communities in natural systems for nutrient source impact assessment. This study provides an improved understanding and metrics for comparing the algal growth response to different nutrient sources.
Prokaryotic gene prediction plays an important role in understanding the biology of organisms and their function with applications in medicine and biotechnology. Although the current gene finders are highly sensitive in finding long genes, their sensitivity decreases noticeably in finding shorter genes (<180 nts). The culprit is insufficient annotated gene data to identify distinguishing features in short open reading frames (ORFs). We develop a deep learning-based method called ProtiGeno, specifically targeting short prokaryotic genes using a protein language model trained on millions of evolved proteins. In systematic large-scale experiments on 4,288 prokaryotic genomes, we demonstrate that ProtiGeno predicts short coding and noncoding genes with higher accuracy and recall than the current state-of-the-art gene finders. We discuss the predictive features of ProtiGeno and possible limitations by visualizing the three-dimensional structure of the predicted short genes. Data, codes, and models are available at https://github.com/tonytu16/protigeno.
Thomas Jun Jewell, Andrew L. Krause, Philip K. Maini
et al.
Throughout developmental biology and ecology, transport can be driven by nonlocal interactions. Examples include cells that migrate based on contact with pseudopodia extended from other cells, and animals that move based on their vision of other animals. Nonlocal integro-PDE models have been used to investigate contact attraction and repulsion in cell populations in 1D. In this paper, we generalise the analysis of pattern formation in such a model from 1D to higher spatial dimensions. Numerical simulations in 2D demonstrate complex behaviour in the model, including spatio-temporal patterns, multi-stability, and the selection of spots or stripes heavily depending on interactions being attractive or repulsive. Through linear stability analysis in $N$ dimensions, we demonstrate how, unlike in local Turing reaction-diffusion models, the capacity for pattern formation fundamentally changes with dimensionality for this nonlocal model. Most notably, pattern formation is possible only in higher than one spatial dimension for both the single species system with repulsive interactions, and the two species system with `run-and-chase' interactions. The latter case may be relevant to zebrafish stripe formation, which has been shown to be driven by run-and-chase dynamics between melanophore and xanthophore pigment cells.
Neurodegenerative disorders (ND) are a group of conditions that affect the neurons in the brain and spinal cord, leading to their degeneration and eventually causing the loss of function in the affected areas. These disorders can be caused by a range of factors, including genetics, environmental factors, and lifestyle choices. Major pathological signs of these diseases are protein misfolding, proteosomal dysfunction, aggregation, inadequate degradation, oxidative stress, free radical formation, mitochondrial dysfunctions, impaired bioenergetics, DNA damage, fragmentation of Golgi apparatus neurons, disruption of axonal transport, dysfunction of neurotrophins (NTFs), neuroinflammatory or neuroimmune processes, and neurohumoral symptoms. According to recent studies, defects or imbalances in gut microbiota can directly lead to neurological disorders through the gut-brain axis. Probiotics in ND are recommended to prevent cognitive dysfunction, which is a major symptom of these diseases. Many in vivo and clinical trials have revealed that probiotics (<i>Lactobacillus acidophilus</i>, <i>Bifidobacterium bifidum</i>, and <i>Lactobacillus casei</i>, etc.) are effective candidates against the progression of ND. It has been proven that the inflammatory process and oxidative stress can be modulated by modifying the gut microbiota with the help of probiotics. As a result, this study provides an overview of the available data, bacterial variety, gut-brain axis defects, and probiotics’ mode of action in averting ND. A literature search on particular sites, including PubMed, Nature, and Springer Link, has identified articles that might be pertinent to this subject. The search contains the following few groups of terms: (1) Neurodegenerative disorders and Probiotics OR (2) Probiotics and Neurodegenerative disorders. The outcomes of this study aid in elucidating the relationship between the effects of probiotics on different neurodegenerative disorders. This systematic review will assist in discovering new treatments in the future, as probiotics are generally safe and cause mild side effects in some cases in the human body.
Vadthya Lokya, Sejal Parmar, Arun K. Pandey
et al.
Abstract In addition to the challenge of meeting global demand for food production, there are increasing concerns about food safety and the need to protect consumer health from the negative effects of foodborne allergies. Certain bio‐molecules (usually proteins) present in food can act as allergens that trigger unusual immunological reactions, with potentially life‐threatening consequences. The relentless working lifestyles of the modern era often incorporate poor eating habits that include readymade prepackaged and processed foods, which contain additives such as peanuts, tree nuts, wheat, and soy‐based products, rather than traditional home cooking. Of the predominant allergenic foods (soybean, wheat, fish, peanut, shellfish, tree nuts, eggs, and milk), peanuts (Arachis hypogaea) are the best characterized source of allergens, followed by tree nuts (Juglans regia, Prunus amygdalus, Corylus avellana, Carya illinoinensis, Anacardium occidentale, Pistacia vera, Bertholletia excels), wheat (Triticum aestivum), soybeans (Glycine max), and kidney beans (Phaseolus vulgaris). The prevalence of food allergies has risen significantly in recent years including chance of accidental exposure to such foods. In contrast, the standards of detection, diagnosis, and cure have not kept pace and unfortunately are often suboptimal. In this review, we mainly focus on the prevalence of allergies associated with peanut, tree nuts, wheat, soybean, and kidney bean, highlighting their physiological properties and functions as well as considering research directions for tailoring allergen gene expression. In particular, we discuss how recent advances in molecular breeding, genetic engineering, and genome editing can be used to develop potential low allergen food crops that protect consumer health.