SEARCH · Search bioRxiv

Results for “Systems Biology”

Search indexed bioRxiv preprints in genomics, neuroscience, cell biology and bioinformatics. Read source abstracts and check manuscript versions; preprints are not peer reviewed.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,423 records · Page 79Linked to original sources

bioRxiv2023

Scalable joint non-negative matrix factorisation for paired single cell gene expression and chromatin accessibility data

Single cell multi-modal technologies provide powerful means to simultaneously profile components of the gene regulatory path-ways of individual cells. These are now being employed to study gene regulatory mechanisms in a variety of biological systems. Tailored computational methods for integration and analysis of these data are much-needed with desirable properties in terms of efficiency -to cope with high dimensionality of the data, inter-pretability -for downstream biological discovery and hypothesis generation, and flexibility -to be able to easily incorporate future modalities. Existing methods cover some but not all of the desirable properties for effective integration of these data. Here we present a highly efficient method, intNMF, for representation and integration of single cell multi-modal data using joint non-negative matrix factorisation which can facilitate discovery of linked regulatory topics in each modality. We provide thorough benchmarking using large publicly available datasets against five popular existing methods. intNMF performs comparably against the current state-of-the-art, and provides advantages in terms of computational efficiency and interpretability of discovered regulatory topics in the original feature space. We illustrate this enhanced interpretability in providing insights into cell state changes associated with Alzheimers disease. int-NMF is available as a Python package with extensive documentation and use-cases at https://github.com/wmorgans/quick_intNMF

bioinformatics↗

bioRxiv2025

Optically detected and radio wave-controlled spin chemistry in cryptochrome

Optically addressable spin systems, such as nitrogen-vacancy centers in diamond, have been widely studied for quantum sensing applications. In this work, we demonstrate that certain flavoproteins -- specifically cryptochrome and iLOV -- which generate spin correlated radical pairs upon optical excitation, also exhibit optically detected magnetic resonance (ODMR). Remarkably, the iLOV protein, commonly used in cellular imaging, displays ODMR contrast approaching 50%. We present initial applications including widefield magnetic field sensing and spatial modulation of photoluminescence using radiofrequency pulses and magnetic field gradients. Our results establish radical pairs in proteins as a novel platform for optically addressable spin systems, offering the key advantages of molecular designability and genetic encodability. Moreover, due to the spin-selective nature of radical pair chemistry, the results lay the groundwork for future radiofrequency-based manipulation of biological systems.

biophysics↗

bioRxiv2026

MetagenomicsBench: A Verifiable Benchmark for Agentic Metagenomics Analysis

Metagenomics has transformed our ability to characterize microbial communities without relying on cultivation, but translating complex microbiome datasets into reliable biological conclusions remains a major analytical challenge. Although AI agents have improved substantially at software engineering and general data analysis, it remains unclear whether they can reliably analyze and interpret real-world metagenomic data. We introduce MetagenomicsBench, a benchmark of 100 verifiable evaluations derived from published datasets spanning community structure, host-microbiome associations, microbial function, longitudinal dynamics, and microbiome interventions. Each evaluation provides experimental data and a deterministic grader that evaluates recovery of a key analytical or biological result. Across 10,200 trajectories from 34 model-harness configurations, the strongest configuration achieved a 60.0% pass rate. Performance varied across functional domains, data modalities, and biological systems, while greater cost, token usage, and tool use did not consistently correspond to higher accuracy. Trajectory analysis showed that agents often produced technically plausible analyses while making mistakes in problem interpretation, statistical reasoning, and biological interpretation. MetagenomicsBench provides a framework for measuring progress toward AI agents that can reliably make and validate the analytical and biological decisions required for metagenomic research.

microbiology↗

bioRxiv2022

Identification of a suitable endogenous control miRNA in bone aging and senescence

ObjectiveMicroRNAs (miRNAs) are promising tools as biomarkers and therapeutic agents in various chronic diseases such as osteoporosis, cancers, type I and II diabetes, and cardiovascular diseases. Considering the rising interest in the regulatory role of miRNAs in bone metabolism, aging, and cellular senescence, accurate normalization of qPCR-based miRNA expression data using an optimal endogenous control becomes crucial. MethodsWe used a systematic approach to select candidate endogenous control miRNAs that exhibit high stability with aging from our miRNA sequence data and literature search. Validation of miRNA expression was performed using qPCR and their comprehensive stability was assessed using the RefFinder tool which is based on four statistical algorithms: GeNorm, NormFinder, BestKeeper, and comparative delta CT. The selected endogenous control was then validated for its stability in mice and human bone tissues, and in bone marrow stromal cells (BMSCs) following induction of senescence and senolytic treatment. Finally, the utility of selected endogenous control versus U6 was tested by using each as a normalizer to measure the expression of miR-34a, a miRNA known to increase with age and senescence. ResultsOur results show that Let-7f did not change across the groups with aging, senescence or senolytic treatment, and was the most stable miRNA, whereas U6 was the least stable. Moreover, using Let-7f as a normalizer resulted in significantly increased expression of miR-34a with aging and senescence and decreased expression following senolytic treatment. However, the expression pattern for miR-34a reversed for each of these conditions when U6 was used as a normalizer. ConclusionsWe show that optimal endogenous control miRNAs, such as Let-7f, are essential for accurate normalization of miRNA expression data to increase the reliability of results and prevent misinterpretation. Moreover, we present a systematic strategy that is transferrable and can easily be used to identify endogenous control miRNAs in other biological systems and conditions.

cell biology↗

bioRxiv2024

ESPRESSO: Spatiotemporal omics based on organelle phenotyping

Omics technologies, including genomics, transcriptomics, proteomics, and metabolomics, have been instrumental to improving our understanding of complex biological systems. Despite fast-pace advancements, a crucial dimension is still left to explore: time. To capture this key parameter, we introduce ESPRESSO (Environmental Sensor Phenotyping RElayed by Subcellular Structures and Organelles), a pioneering technique that provides high-dimensional phenotyping resolved in space and time. Through a novel paradigm, ESPRESSO combines fluorescent labeling, advanced microscopy and bioimage and data analysis to extract morphological and functional information of the organelle network unveiling phenotypic changes over time at the single-cell level. In this work, we present ESPRESSOs methodology and its application across numerous cellular systems, showcasing its ability to discern cell types, stress response, differentiation and immune cells polarization. We then correlate ESPRESSO phenotypic changes with gene expression and demonstrate its applicability to 3D cultures, offering a path to revolutionizing biological exploration, providing invaluable insights into cellular states in both space and time.

bioengineering↗

bioRxiv2025

Accurate prediction of thermoresponsive phase behavior of disordered proteins

Protein responses to environmental stress, particularly temperature fluctuations, have long been a subject of investigation, with a focus on how proteins maintain homeostasis and exhibit thermoresponsive properties. While UCST-type (upper critical solution temperature) phase behavior has been studied extensively and can now be predicted reliably using computational models, LCST-type (lower critical solution temperature) phase transitions remain less explored, with a lack of computational models capable of accurate prediction. This gap limits our ability to probe fully how proteins undergo phase transitions in response to temperature changes. Here, we introduce Mpipi-T, a residue-level coarse-grained model designed to predict LCST-type phase behavior of proteins. Parametrized using both atomistic simulations and experimental data, Mpipi-T accounts for entropically driven protein phase separation that occurs upon heating. Accordingly, Mpipi-T predicts temperature-driven protein behavior quantitatively in both single- and multi-chain systems. Beyond its predictive capabilities, we demonstrate that Mpipi-T provides a framework for uncovering the molecular mechanisms underlying heat stress responses, offering new insights into how proteins sense and adapt to thermal changes in biological systems.

biophysics↗

bioRxiv2025

CellFlow enables generative single-cell phenotype modeling with flow matching

High-content phenotypic screens provide a powerful strategy for studying biological systems, but the scale of possible perturbations and cell states makes exhaustive experiments unfeasible. Computational models that are trained on existing data and extrapolate to correctly predict outcomes in unseen contexts have the potential to accelerate biological discovery. Here, we present CellFlow, a flexible framework based on flow matching that can model single cell phenotypes induced by complex perturbations. We apply CellFlow to various phenotypic screens, accurately predicting expression responses to a wide range of perturbations, including cytokine stimulation, drug treatments and gene knockouts. CellFlow successfully modeled developmental perturbations at the whole-embryo scale and guided cell fate and organoid engineering by predicting heterogeneous cell populations arising from combinatorial morphogen treatments and by performing a virtual organoid protocol screen. Taken together, CellFlow has the potential to accelerate discovery from phenotypic screens by learning from existing data and generating phenotypes induced by unseen conditions.

bioinformatics↗

bioRxiv2025

Rapid degradation of 6 class I HDAC complexes reveals minimal functional overlap between complexes

The class 1 HDACs 1, 2 and 3 form seven families of distinct large multiprotein complexes that regulate gene expression via deacetylation of lysines in histone tails. The degree of redundancy and functional overlap between complexes and their primary gene targets, remains unknown. We used CRISPR/Cas9 to independently tag HDAC complexes with FKBP12F36V in HCT116 cells enabling rapid (<1 hr), PROTAC-mediated, degradation. RNA sequencing at 6 h reveals that together, the 4 major complexes (CoREST, NuRD, NCoR/SMRT and SIN3A) perturbed >50% of expressed genes. More than 60% of these are specific to an individual complex. Of genes regulated by more than one complex, approaching 50% are reciprocally regulated such that HDAC complexes act as antagonistic regulators. Homer analysis strongly suggests that the complexes are reliant on different transcription factors. This is the first study to identify the primary targets of individual HDAC complexes and directly compare the effects of rapid degradation on gene regulation in the same biological system.

molecular biology↗

bioRxiv2025

Flexible navigation with neuromodulated cognitive maps

Animals develop specialized cognitive maps during navigation, constructing environmental representations that facilitate efficient exploration and goal-directed planning. The hippocampal CA1 region is implicated as the primary neural substrate for cognitive mapping, housing spatially tuned cells that adapt based on behavioral patterns and internal states. Computational approaches to modeling these biological systems have employed various methodologies. Although labeled graphs with local spatial information and deep neural networks have provided computational frameworks for spatial navigation, significant limitations persist in modeling one-shot adaptive mapping. We introduce a biologically inspired place cell architecture that develops cognitive maps during exploration of novel environments. Our model implements a simulated agent for reward-driven navigation that forms spatial representations online. The architecture incorporates behaviorally relevant information through neuromodulatory signals that respond to environmental boundaries and reward locations. Learning combines rapid Hebbian plasticity, lateral competition, and targeted modulation of place cells. Analysis of the capabilities of the model on a variety of environments demonstrates our approachs efficiency, achieving in one shot what traditional RL models require thousands of epochs to learn. The simulation results show that the agent successfully explores and navigates to the target locations in various environments, showing adaptability when the reward positions change. Analysis of neuromodulated place cells reveals dynamic changes in neuronal density and tuning field size after behaviorally significant events. These findings align with experimental observations of reward effects on hippocampal spatial cells while providing computational support for the efficacy of biologically inspired approaches to cognitive mapping.

neuroscience↗

bioRxiv2025

scCotag: Diagonal integration of single-cell multi-omics data via prior-informed co-optimal transport and regularized barycentric mapping

Recent advances in high-throughput single-cell technologies have enabled characterization of cellular states across distinct omics layers, yielding complementary insights into the organization of biological systems. To jointly leverage these disparate modalities, computational methods have been developed to align non-overlapping cell populations profiled in different omics layers, a task known as diagonal integration, thereby facilitating more fine-grained biological interpretations. However, existing integration approaches unrealistically assume that all cells are alignable and treat prior-knowledge-derived feature correspondences as fully reliable, retaining them unrefined throughout the alignment process. In this work, we introduce scCotag, a co-optimal transport (COOT)-based deep learning frame-work for diagonal integration of single-cell multi-omics data, to address these limitations. scCotag first infers cell alignment and feature correspondence with prior-informed COOT in an iterative manner, and then leverages the resulting transport plans to jointly learn cell and feature embeddings via regularized barycentric mapping and graph reweighting. Systematic benchmarking on human brain, bone marrow, and blood single-cell RNA-seq and ATAC-seq datasets demonstrates the overall superior performance of scCotag over state-of-the-art methods in both cell alignment accuracy and embedding accuracy. In particular, scCotag excels in simulated imbalance scenarios where not all cells are alignable. Applying scCotag to data in postmortem brains of Alzheimers disease (AD) and non-AD (NoAD) patients further yields more fine-grained biological interpretation of regulatory mechanisms. Together, these results demonstrate that scCotag provides a robust framework for single-cell diagonal integration and regulatory inference.

bioinformatics↗

bioRxiv2026

Carbon Capture Modeling and Simulation Platform: A Coupled Microalgal Bioreactor-Yeast Fermentation Approach for Bioethanol

This article presents the development of an advanced modeling and simulation platform for carbon capture systems, with a focus on integrated process analysis from upstream CO2 capture through to bioethanol production. The platform supports the evaluation of CO2 mitigation technology by coupling mathematical bioprocess models with an interactive desktop application. The biological system employs Chlorella vulgaris microalgae to fix CO2 through photosynthesis and generate carbohydrate substrates, which are subsequently converted to bioethanol by Saccharomyces cerevisiae yeast via fermentation. The simulation integrates three established kinetic models--the Monod, Logistic, and Luedeking-Piret models--to predict biomass growth, substrate consumption, and ethanol yield under varying operational conditions. A closed-loop CO2 recycling subsystem captures fermentation off-gases and reintroduces them into the bioreactor, enhancing overall carbon utilization efficiency. Three representative simulation scenarios demonstrated process efficiencies ranging from 1.09% to 93.78% of the theoretical maximum CO2-to-ethanol conversion efficiency, confirming the platforms capacity to evaluate a wide operational envelope. The Electron/React-based desktop application provides real-time visualization, interactive 3D bioreactor models, and a simulation history module, making it accessible to researchers, engineers, and students. The platform serves as a digital twin that bridges rigorous bioprocess mathematics with intuitive user interaction, providing a cost-effective tool for designing and optimizing sustainable carbon capture and biofuel production systems.

bioengineering↗

bioRxiv2019

Living materials with programmable functionalities grown from engineered microbial co-cultures

Biological systems assemble tissues and structures with advanced properties in ways that cannot be achieved by man-made materials. Living materials self-assemble under mild conditions, are autonomously patterned, can self-repair and sense and respond to their environment. Inspired by this, the field of engineered living materials (ELMs) aims to use genetically-engineered organisms to generate novel materials. Bacterial cellulose (BC) is a biological material with impressive physical properties and low cost of production that is an attractive substrate for ELMs. Inspired by how plants build materials from tissues with specialist cells we here developed a system for making novel BC-based ELMs by addition of engineered yeast programmed to add functional traits to a cellulose matrix. This is achieved via a synthetic symbiotic culture of bacteria and yeast (Syn-SCOBY) approach that uses a stable co-culture of Saccharomyces cerevisiae with BC-producing Komagataeibacter rhaeticus bacetria. Our Syn-SCOBY approach allows inoculation of engineered cells into simple growth media, and under mild conditions materials self-assemble with genetically-programmable functional properties in days. We show that co-cultured yeast can be engineered to secrete enzymes into BC, generating autonomously grown catalytic materials and enabling DNA-encoded modification of BC bulk material properties. We further developed a method for incorporating S. cerevisiae within the growing cellulose matrix, creating living materials that can sense chemical and optical inputs. This enabled growth of living sensor materials that can detect and respond to environmental pollutants, as well as living films that grow images based on projected patterns. This novel and robust Syn-SCOBY system empowers the sustainable production of BC-based ELMs.

synthetic biology↗

bioRxiv2019

Isolated occurrences of membrane perturbation by mechanosensing from weakly aggregating silver nanoparticles

Silver nanoparticles (AgNPs) have wide-ranging applications, including as additives in consumer products and in medical diagnostics and therapy. Therefore understanding how AgNPs interact with biological systems is important for ascertaining any potential health risks due to the likelihood of high levels of human exposure. Besides any severe, acute effects, it is desirable to understand more subtle interactions that could lead to milder, chronic health impacts. Nanoparticles are small enough to be able to enter biological cells and interfere with their internal biochemistry. The initial contact between nanoparticle and cell is at the plasma membrane. To gain fundamental mechanistic insight into AgNP-membrane interactions, we investigate these phenomena in minimal model systems using a wide-range of biophysical approaches applied to lipid vesicles. We find a strong dependence on the medium composition, where colloidally stable AgNPs in a glucose buffer have negligible effect on the membrane. However, at a physiological salt concentrations, the AgNPs start to weakly aggregate and sporadic but significant membrane perturbation events are observed. Under these latter conditions, transient poration and structural remodelling of some vesicle membranes is observed. We observe that the fluidity of giant vesicle membranes universally decreases by an average of 16% across all vesicles. However, we observe a small population of vesicles display a significant change in mechanical properties with lower bending rigidity and higher membrane tension. Therefore we argue that the isolated occurrences of membrane perturbation by AgNPs are due to low probability mechanosensing events of AgNP aggregation at the membrane.\n\nGRAPHICAL ABSTRACT\n\nO_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=109 SRC=\"FIGDIR/small/623678v1_ufig1.gif\" ALT=\"Figure 1\">\nView larger version (54K):\norg.highwire.dtl.DTLVardef@196aa51org.highwire.dtl.DTLVardef@87c12borg.highwire.dtl.DTLVardef@9b2096org.highwire.dtl.DTLVardef@7877f6_HPS_FORMAT_FIGEXP M_FIG C_FIG

biophysics↗

bioRxiv2025

ParTIpy: A Scalable Framework for Archetypal Analysis and Pareto Task Inference

MotivationTrade-offs between different functions or tasks are pervasive across scales in biological systems. For example, individual cells cannot perform all possible functions simultaneously; instead they allocate limited resources to specialize in subsets of tasks by activating specific gene expression programs. Pareto Task Inference (ParTI) is a framework for analyzing biological trade-offs grounded in the theory of multi-objective optimality. However, existing software implementations of ParTI lack scalability to large datasets and do not integrate well with standard biological data analysis workflows, especially in the context of single-cell transcriptomics, limiting broader adoption. ResultsWe have developed ParTIpy (Pareto Task Inference in Python), an open-source Python package that combines advances in optimization and coreset methods to scale archetypal analysis, the primary algorithm underlying ParTI, to large-scale datasets. By providing additional tools to characterize archetypes, comprehensive documentation, and adopting standard scverse data structures, ParTIpy facilitates seamless integration into existing analysis workflows and broadens accessibility, particularly within the single-cell community. We demonstrate how ParTIpy can be used to study intra-cell-type gene expression variability through the lens of task allocation, offering a principled alternative to methods that impose discrete cell state classifications on inherently continuous variation. Availability and implementationParTIpys open-source code is available on GitHub (https://github.com/saezlab/ParTIpy) and pypi (https://pypi.org/project/partipy). Documentation is available at https://partipy.readthedocs.io. The code to reproduce the results of this paper is on GitHub (https://github.com/saezlab/ParTIpy_paper)

bioinformatics↗

bioRxiv2025

System-level health profiling from blood DNA methylation with explainable deep learning

Genome-scale DNA methylation (DNAm) profiles capture organismal physiology, but most predictive models lack transparency and multi-level applicability. Here we develop an explainable framework that quantifies respiratory, cardiovascular, and metabolic status as bounded health scores (0-1) derived from sex-specific clinical reference ranges and disease penalties, and then predicts these scores from whole-blood DNAm. Using Generation Scotland and case-control samples (n = 14,496 individuals), we screened 39 covariates for disease relevance and DNAm predictability, yielding system- relevant panels that were aggregated into scores. We compressed DNAm profiles with a protein-interaction-guided autoencoder, and trained health predictors on 128- dimensional embeddings using fully connected networks. On held-out samples, models reproduced the composite scores with strong rank agreement (Spearman {rho} = 0.87, R2 = 0.71 for respiratory health; {rho} = 0.82, R2 = 0.66 for cardiovascular; {rho} = 0.81, R2 = 0.64 for metabolic) and recover expected population structure in a generally healthy cohort, with clear separation between "single-system low" and "multi-system low" phenotypes, and graded coupling across systems without redundancy. Further, the top features retrieved from each explainable predictor aligned with system biology: airway epithelial repair, hypoxia and inflammatory trafficking for respiratory; endothelial remodeling and cardiomyocyte programs for cardiovascular; glucose-lipid metabolism and metaflammation for metabolic. These results show that DNAm embeddings can yield accurate, transparent, and system-aware health profiling from blood, providing actionable summaries while revealing the molecular processes the models use to infer multi-system status. This approach positions DNAm embeddings plus interpretable penalty targets as a practical bridge from epigenomic signal to system-level triage and is extensible for evaluation in larger, more diverse cohorts.

bioinformatics↗

bioRxiv2018

Differences between the de novo proteome and its non-functional precursor can result from neutral constraints on its birth process, not necessarily from natural selection alone

Proteins are among the most important constituents of biological systems. Because all proteins ultimately evolved from previously non-coding DNA, the properties of these non-coding sequences and how they shape the birth of novel proteins are also expected to influence the organization of biological networks. When trying to explain and predict the properties of novel proteins, it is of particular importance to distinguish the contributions of natural selection and other evolutionary forces. Studies in the field typically use non-coding DNA and GC-content-based random-sequence models to generate random expectations for the properties of novel functional proteins. Deviations from these expectations have been interpreted as the result of natural selection. However, interpreting such deviations requires a yet-unattained understanding of the raw material of de novo gene birth and its relation to novel functional proteins. We mathematically show how the importance of the \"junk\" polypeptides that make up this raw material goes beyond their average properties and their filtering by natural selection. We find that the mean of any property among novel functional proteins also depends on its variance among junk polypeptides and its correlation with their rate of evolutionary turnover. In order to exemplify the use of our general theoretical results, we combine them with a simple model that predicts the means and variances of the properties of junk polypeptides from the genomic GC content alone. Under this model, we predict the effect of GC content on the mean length and mean intrinsic disorder of novel functional proteins as a function of evolutionary parameters. We use these predictions to formulate new evolutionary interpretations of published data on the length and intrinsic disorder of novel functional proteins. This work provides a theoretical framework that can serve as a guide for the prediction and interpretation of past and future results in the study of novel proteins and their properties under various evolutionary models. Our results provide the foundation for a better understanding of the properties of cellular networks through the evolutionary origin of their components.

evolutionary biology↗

bioRxiv2022

Engineering ssRNA tile filaments for (dis)assembly and membrane binding

Cytoskeletal protein filaments such as actin and microtubules confer mechanical support to cells and facilitate many cellular functions such as motility and division. Recent years have witnessed the development of a variety of molecular scaffolds that mimic such cytoskeletal filaments. Indeed, filaments that are programmable and compatible with biological systems may prove useful in studying or substituting such proteins. Here, we explore the use of ssRNA tiles to build and modify cytoskeletal-like filaments in vitro. We engineer a number of functionalities that are crucial to the function of natural cytoskeletal systems into the ssRNA tiles, including the abilities to assemble or disassemble filaments, to tune the filament stiffness, to induce membrane binding, and to attach binding proteins. The work paves the way for building dynamic cell-like cytoskeletal systems made out of rationally designed ssRNA tiles that can be transcribed in natural or synthetic cells.

synthetic biology↗

bioRxiv2025

A Critical Re-evaluation of the Somatic Mutation Theory (SMT)Conceptual and Quantitative Limits of the Cell-Centric Paradigm

The Somatic Mutation Theory (SMT) frames cancer as a stochastic process driven by the accumulation of random mutations in somatic cells. Using a Poisson-Erlang waiting-time model, we test whether empirical mutation rates and cell-division frequencies permit such multi-hit carcinogenesis within biologically realistic timescales. Even under optimistic assumptions, expected waiting times exceed progenitor-cell lifespans by several orders of magnitude, rendering sequential multi-hit carcinogenesis statistically and biologically implausible. The analysis exposes SMTs internal inconsistencies and supports a paradigm shift from mutation accumulation to communication breakdown as the root cause of cancer. Companion Paper NoticeThis preprint constitutes Part I of a coordinated two-part series on the origins of cancer. Part I critically re-examines the Somatic Mutation Theory (SMT), demonstrating its conceptual and quantitative limits as a cell-centric paradigm. Part II - From Cell-Centric to System-Centric Carcinogenesis: The Repair and Capacity Adaptation (RCA) Conceptual Framework (Olsen & Liisberg, 2025b; Zenodo 10.5281/zenodo.17495975) - presents the complementary systems-biology framework that explains carcinogenesis as failure of progenitor-cell maturation within a degraded connective-tissue microenvironment. Together, the two papers mark a paradigm shift from mutation accumulation to communication-system failure as the root cause of cancer. A laymans summary of Part I is available as supplementary material for non-specialist readers.

cancer biology↗