The impetus behind these fellowships lies in the rapidly expanding availability of large-scale biological datasets and the increasing necessity for sophisticated computational frameworks to interpret them. Yung S. Lie, PhD, President and CEO of the Damon Runyon Cancer Research Foundation, emphasizes the crucial role of computational expertise in precision medicine, where modeling and data integration are vital for dissecting tumor heterogeneity and treatment responses. The selected fellows epitomize this interdisciplinary approach, bridging “dry” lab quantitative sciences with “wet” lab biological insights to pioneer novel avenues in cancer understanding and intervention.

Among the fellowship recipients is Dr. Minsoo Kim, who focuses on the enigmatic presence of aneuploid cells—cells with abnormal chromosome numbers—in ostensibly healthy breast tissue. Challenging long-held assumptions that normal cells uniformly maintain chromosomal integrity, Dr. Kim’s research investigates these rare aneuploid populations as potential early harbingers of breast cancer. By developing a heterogeneous graph neural network (GNN), his work will jointly model single-cell copy number variations and gene expression data, representing genes, cells, and chromosome segments as distinct nodes. This nuanced modeling approach aims to disentangle gene expression changes driven by chromosomal gains or losses from other transcriptional variations.

Crucially, Dr. Kim intends to extend this computational framework into spatial transcriptomics, which retains the spatial context of gene expression within tissue architecture. This enhancement is designed to illuminate how the microenvironment influences aneuploid cell behavior and interactions, potentially revealing biomarkers for early detection and mechanisms of cancer risk stratification. By applying these analyses to longitudinal breast tissue samples from patients monitored over years, where some subsequently developed cancer, the project aspires to not only refine predictive diagnostics but also offer clinicians tools for earlier, more targeted intervention strategies.

Dr. Sahana Kuthyar’s research addresses a pressing clinical challenge: the elevated risk of severe lung infections in cancer patients undergoing immunosuppressive therapies like chemotherapy and radiation. These treatments, while efficacious against tumors, impair myeloid immune components critical for combating bacterial pathogens, leaving patients vulnerable to conditions such as pneumonia. Moreover, the common clinical practice of providing supplemental oxygen further complicates this risk by altering the pulmonary environment to favor aggressive bacterial proliferation. Dr. Kuthyar’s investigation bridges human and murine models to unravel this complex interplay.

Her computational strategy leverages hierarchical network modeling to integrate gene expression profiles with metabolomic data, applying multi-omics factor analysis for a holistic view of microbial and host immune dynamics under hyperoxic conditions. By cross-validating predictive models between human patients and mouse models, the study aims to iteratively refine understanding of how bacterial adaptation and immune suppression converge to create critical infection vulnerabilities. The insights garnered here may pave the way for predictive diagnostics and novel therapeutic approaches to mitigate life-threatening infections in immunocompromised cancer populations.

Matthew Leventhal, PhD, embarks on a pioneering inquiry into sex chromosome biology within cancer, focusing on the differential roles of active and inactive X chromosomes in females—a subject deeply intertwined with oncogenic potential. Given that females carry two X chromosomes with one subjected to early developmental silencing, mutations impacting the active X chromosome may have outsized consequences on cellular function and tumor progression. Dr. Leventhal’s work centers on developing computational tools capable of resolving the haplotype-specific copy number of chromosomes from bulk whole-genome sequencing data, correcting phasing errors that have historically obscured distinctions between active and inactive X chromosome alterations.

Integrating DNA sequencing with RNA-seq expression data, this methodology will allow for the first pan-cancer analysis of X chromosome dynamics across more than 8,500 tumors spanning 31 cancer types. The goal is to identify recurrent copy number alterations preferentially affecting either the active or inactive X, potentially uncovering novel oncogenic drivers or vulnerabilities previously masked due to analytical limitations. Additionally, determining whether such chromosomal alterations exist in precancerous cells could have transformative implications for early detection and intervention strategies tailored to sex chromosome biology.

The innovations promised by these fellows are testament to the evolving landscape of cancer research, where computational advancements are indispensable to dissecting biological complexity. The utilization of graph neural networks, multi-omics integration, and sophisticated haplotype phasing models exemplifies the next frontier of oncological inquiry, promising heightened precision in diagnosis, prognosis, and treatment. Beyond their individual research agendas, these scientists exemplify the Damon Runyon Foundation’s vision of cultivating interdisciplinary talent equipped to unravel cancer’s multifaceted biology.

Since 1946, the Damon Runyon Cancer Research Foundation has championed early-career investigators, recognizing that the initial years of scientific pursuit are critical for unleashing transformative discoveries. Over $491 million invested and nearly 4,100 funded scientists reflect an enduring commitment to nurturing high-risk, high-reward research. The foundation’s outstanding track record, highlighted by thirteen Nobel laureates among its alumni, underscores its impact on the global cancer research community.

These current fellowships reinforce the need to blur conventional boundaries between computational and biological sciences, reinforcing a paradigm where machine learning algorithms and spatial data are indispensable complements to experimental biology. As the biological sciences grapple with data of unprecedented scale and complexity, the fusion of quantitative expertise and biological insight will catalyze breakthroughs in understanding cancer’s origins, progression, and treatment resistance.

The relevance of this fellow-supported research extends to personalized and precision medicine, where patient-specific molecular data can guide tailored therapeutic regimens. Detecting early aneuploid cell populations, predicting infection risks in susceptible patients, and elucidating sex chromosome influences represent concrete ways in which computational biology is reshaping cancer care. Through these fellowships, the Damon Runyon Foundation equips young scientists with not only resources but also mentorship from leaders in computational and biological cancer research, creating a fertile environment for interdisciplinary innovation.

As these fellows progress, their work is poised to impact fundamental understanding and clinical strategies alike. Whether refining early detection algorithms for breast cancer, unearthing microbial-immune crosstalk in cancer-associated pneumonia, or decoding X chromosome alterations across cancers, these efforts embody a new wave of cancer research empowered by computational sophistication. The field awaits the ripple effects of their discoveries as they translate complex biological data into actionable knowledge with the potential to save lives.

In sum, the 2026 Damon Runyon Quantitative Biology Fellows symbolize a convergence of technology and biology at a pivotal moment in cancer research. Their ambitious projects harness state-of-the-art computational methodologies to tackle profound questions about cancer initiation, progression, and patient vulnerability. Supported by visionary funding and mentorship, these scholars exemplify the future of biomedical research, where multidisciplinary collaboration and quantitative prowess unlock mysteries once deemed impenetrable.

Subject of Research: Computational approaches to cancer biology focusing on early detection, infection risk in immunocompromised patients, and sex chromosome genomics in cancer.

Article Title: Unlocking Cancer’s Complexities: How Computational Pioneers are Shaping the Future of Oncology

News Publication Date: 2026

Web References: http://damonrunyon.org/

Keywords: cancer research, computational biology, machine learning, graph neural networks, spatial transcriptomics, multi-omics analysis, cancer immunology, X chromosome, aneuploidy, precision medicine, early cancer detection, network modeling

Traditionally, biologists have been constrained by analytical techniques that sacrifice spatial information, reducing tissues to a mixture of cells where location is lost. This limitation has hindered our comprehension of how cellular neighborhoods influence physiological and pathological states. The emergence of spatial transcriptomics and proteomics now permits high-resolution mapping of RNA and protein molecules inside intact tissues, lending an unprecedented spatial dimension to molecular profiling. Spatial transcriptomics reveals the location-specific gene expression patterns, whereas spatial proteomics identifies the distribution and interplay of functional proteins, thereby elucidating the molecular choreography underlying cellular behavior.

IRB Barcelona’s new platform uniquely integrates multiple core technologies encompassing spatial genomics, proteomics, histopathology, advanced microscopy, and bioinformatics into a seamless workflow designed to generate comprehensive spatially resolved molecular maps. This integrated approach not only enables rigorous sample processing and data acquisition but also incorporates sophisticated computational tools to interpret multilayered datasets. By combining these modalities, researchers can create detailed molecular atlases that reveal how distinct cell types and molecular states coalesce to maintain tissue homeostasis or drive disease progression.

The launch of this platform reflects IRB Barcelona’s longstanding commitment to pioneering technologies that push the boundaries of molecular biology. Over the last two decades, the institute has been a trailblazer in genomic microarrays and single-cell gene expression profiling from minimal samples, establishing itself as a reference center of excellence. Their prior innovations in proteomics, including advanced top-down analysis techniques, and the adoption of light-sheet microscopy for three-dimensional tissue imaging, have laid the foundation for this next leap forward into spatial biology.

This powerful platform facilitates detailed investigation of a wide array of diseases characterized by complex tissue architecture, including cancer, neurodegenerative disorders, cardiovascular ailments, and immune dysfunction. For instance, in oncology, spatial omics can elucidate the cellular heterogeneity within tumors, map the spatial distribution of resistant cell subpopulations, and unravel cellular interactions that influence tumor microenvironment and therapy response. Such spatially-informed molecular data are critical for understanding why certain therapies fail and for identifying novel, spatially targeted therapeutic interventions.

The uniqueness of IRB Barcelona’s initiative lies not only in its technological sophistication but also in its multidisciplinary and collaborative framework. By coordinating expertise from multiple core facilities, the platform delivers an end-to-end pipeline that spans from sample preparation to deep computational analysis. This holistic integration ensures scientific robustness, reproducibility, and the generation of high-resolution spatial datasets that can be cross-compared across studies and over time, accelerating discovery and translational applications.

Moreover, this platform serves as a national hub and a collaborative nexus, opening its infrastructure to the wider scientific community, including academic institutions, hospitals, and industry collaborators. Such open access fosters synergy, drives innovation, and broadens the impact of spatial omics technologies across Spain and internationally. It is envisaged that this initiative will significantly propel precision medicine, enabling patient-specific molecular diagnostics and the development of personalized therapeutic strategies grounded in spatial cellular biology.

A critical aspect of this platform is its integration of advanced computational methods. Spatial omics generates complex, multilayered data that requires novel bioinformatics algorithms to align and co-analyze transcriptomic, proteomic, and phenotypic information within spatial coordinates. IRB Barcelona’s bioinformatics teams are developing and implementing these sophisticated pipelines to construct multidimensional molecular landscapes of tissues. Such atlases not only enhance our understanding of tissue organization and function but also provide invaluable resources for hypothesis generation and mechanistic studies.

The platform is also a testament to successful collaborative funding efforts, having been supported by Spanish and Catalan governmental bodies, Next Generation funds, and prominent foundations such as the Spanish Association Against Cancer, La Caixa Foundation, and the BBVA Foundation. This financial backing underscores the importance and potential impact of spatial omics on public health and biomedical research.

Looking ahead, the integration of spatial omics with other emerging technologies such as single-cell multi-omics and advanced imaging modalities promises to unlock even deeper insights into cellular ecosystems. The ability to spatially resolve multiple biomolecular layers simultaneously will provide a holistic view of biological systems, bridging the gap between molecular detail and tissue physiology. This comprehensive understanding is essential to confront the complexities of human diseases and to develop innovative treatment paradigms.

By enabling researchers to ‘see biology in place’, IRB Barcelona’s Spatial Omics Platform is not merely an incremental technological upgrade but represents a paradigm shift in life sciences. It turns the metaphor of the body as a city into a tangible reality, where cells, genes, and proteins are mapped with neighborhood precision. This spatial perspective is critical for decoding cellular behavior within the rich tapestry of tissue architecture and microenvironmental influences, ultimately advancing both basic biology and precision medicine.

In sum, this pioneering facility positions IRB Barcelona at the forefront of spatial biology, empowering scientists to unlock the spatial dimension of molecular biology that has remained elusive until now. The resulting knowledge is expected to transform our approach to diagnosing, treating, and preventing diseases with unprecedented accuracy and specificity, heralding a new era in biomedical research.

Subject of Research: Spatial Omics, Spatial Transcriptomics, Spatial Proteomics, Integrated Molecular Profiling

Article Title: IRB Barcelona Launches Spain’s First Integrated Spatial Omics Platform Revolutionizing Molecular Mapping in Tissues

News Publication Date: 9 February 2026

Image Credits: IRB Barcelona

Keywords: Genomics, Proteomics, Microscopy, Cancer, Bioinformatics, Health and Medicine

The basement membrane is a pivotal structure in the body that provides support and anchorage for various cell types, playing a crucial role in tissue architecture and function. In glioblastoma, alterations in the basement membrane have been implicated in tumor progression, invasiveness, and patient prognosis. By employing advanced machine learning techniques, Liu and colleagues were able to sift through vast amounts of transcriptomic data to identify gene signatures that are closely linked to the basement membrane’s characteristics in glioblastoma tissues.

The study utilized cutting-edge single-cell RNA sequencing, a technique that allows researchers to analyze gene expression at a single-cell resolution. This approach is revolutionary as it reveals the heterogeneity present within tumors, providing insights into the various cell types involved in tumor growth and invasiveness. Previous studies had primarily focused on bulk tissue analysis, often obscuring the diversity of individual cells. This granular view offered by single-cell sequencing has enabled the identification of specific cell populations that may play decisive roles in glioblastoma biology.

Spatial transcriptomics further enriches our understanding by retaining the spatial context of gene expression within tissue samples. By mapping gene activity back to their original location in the tissue, researchers can observe the interactions between tumor cells and their surrounding microenvironment. Liu et al. effectively combined these techniques to create a comprehensive portrait of glioblastoma, resulting in the identification of genes that not only characterize the cancer but also implicate the basement membrane’s role in tumor behavior.

The researchers applied machine learning algorithms to analyze the data obtained from these advanced techniques. This computational approach enhanced their ability to discern patterns and relationships within the data that may not be immediately apparent through traditional analytical strategies. By training models on the transcriptomic profiles of glioblastoma samples, they could predict the relevance of specific genes related to the basement membrane, leading to the discovery of a novel gene signature.

Significantly, the identified gene signature holds promise not only for understanding glioblastoma pathology but also for potential therapeutic applications. Targeting the basement membrane-related pathways that are disrupted in glioblastoma may represent a novel strategy for treatment. This is particularly crucial given the limited effectiveness of current therapies, which often fail to address the aggressive nature of this malignancy and the challenges posed by the tumor microenvironment.

An intriguing aspect of the research is its potential to guide personalized medicine in neuro-oncology. By characterizing tumors based on their genetic signatures, clinicians may be able to tailor treatment plans that are more aligned with a patient’s unique tumor profile. The implications of this study extend to prognostic assessments as well, providing insights into which patients might have a more favorable or unfavorable outcome based on the expression of specific genes associated with the basement membrane.

In addition to the clinical implications, this research exemplifies the transformative power of interdisciplinary approaches in science. The fusion of machine learning with molecular biology and spatial analysis underscores how advanced computational methods can enhance our comprehension of complex biological systems. As scientists continue to explore the intersections of technology and medicine, innovations like those presented by Liu et al. will likely catalyze further breakthroughs in cancer research.

This research also highlights the importance of collaboration and resource-sharing within the scientific community. By utilizing publicly available datasets and encouraging open access to methodologies, researchers can build upon each other’s work, accelerating the pace of discovery. The transparent sharing of data and techniques fosters an environment where collective knowledge can flourish, leading to faster advancements in understanding and treating diseases like glioblastoma.

As we digest the findings from Liu et al.’s research, it is essential to recognize the broader implications for the field of cancer research. The methodologies applied in this study are not limited to glioblastoma; they can be adapted to investigate other malignancies and complex diseases. This adaptability underscores the versatility of machine learning and advanced transcriptomic techniques in unveiling the molecular underpinnings of various health conditions.

Moreover, as the field progresses, it’s crucial to consider the ethical implications of using machine learning in healthcare. Ensuring that patient data is handled with the utmost care and maintaining privacy standards will be critical as research becomes increasingly reliant on large datasets. Adopting guidelines for ethical research practices will be necessary to build public trust and ensure responsible use of innovative technologies in medicine.

Looking ahead, the next steps following this pivotal research will involve clinical trials to validate the utility of the identified gene signature in a therapeutic context. It will be critical to determine how these findings can translate into tangible benefits for patients with glioblastoma. This may involve developing targeted therapies that can effectively modulate the functions of the disrupted basement membrane pathways identified in this study.

In conclusion, Liu and colleagues have made a significant stride in uncovering the genetic signatures associated with glioblastoma through the integration of machine learning, single-cell RNA sequencing, and spatial transcriptomics. Their work not only elucidates the complexities of tumor biology but also paves the way for future research that might lead to novel therapeutic avenues. As this field continues to evolve, the collaboration of computational and biological sciences will remain at the forefront of uncovering solutions for one of oncology’s most challenging adversaries.

Ultimately, the discovery of a basement membrane-related gene signature in glioblastoma not only contributes to our understanding of tumor biology but also ignites hope for improved patient outcomes through personalized therapies. This remarkable intersection of technology and medicine epitomizes the future of cancer treatment, where data-driven insights will guide innovative interventions tailored to the individual characteristics of each patient’s tumor.

Subject of Research: Glioblastoma and basement membrane-related gene signatures

Article Title: Machine learning-enhanced discovery of a basement membrane-related gene signature in glioblastoma via single-cell and spatial transcriptomics.

Article References: Liu, Z., Yang, Y., Fang, H. et al. Machine learning-enhanced discovery of a basement membrane-related gene signature in glioblastoma via single-cell and Spatial transcriptomics. J Transl Med 23, 1325 (2025). https://doi.org/10.1186/s12967-025-06918-0

Image Credits: AI Generated

DOI: https://doi.org/10.1186/s12967-025-06918-0

Keywords: Glioblastoma, basement membrane, machine learning, single-cell transcriptomics, spatial transcriptomics, gene signature, cancer research.

Despite the substantial progress in applying SSL to scRNA-seq data, a significant gap remains in understanding the transferability of these pretrained models to other related fields, particularly spatial transcriptomics. Spatial transcriptomics, which adds a spatial dimension to gene expression profiles, holds the potential to revolutionize how we understand cellular environments and interactions within tissues. However, the extent to which models pretrained on scRNA-seq data can be adapted for spatial transcriptomics has not been rigorously explored until now.

In their study, the authors meticulously evaluated three distinct SSL models: a random mask strategy, a gene programme mask, and Barlow Twins. Each model was pretrained on scRNA-seq data and subjected to various assessments using spatial transcriptomics datasets with a focus on cell-type prediction and spatial clustering. The findings revealed that the random mask strategy SSL model outperformed its counterparts, indicating a significant potential for this approach in spatially mapping cellular information.

The study unraveled an intriguing facet of the research—models trained from scratch on spatial transcriptomics data exhibited superior performance compared to fine-tuned SSL models in the context of cell-type prediction. This discrepancy raises critical questions about the underlying differences in data characteristics between scRNA-seq and spatial transcriptomics, prompting further exploration into the reasons behind this phenomenon. Understanding these potential domain gaps could provide valuable insights into the development of more effective models that can bridge the divide between these two domains.

Moreover, the researchers delved into the impact of data processing techniques on model performance. Their analysis of multiple imputation methods and scenarios of data degradation spotlighted the complexities affiliated with gene imputation processes, revealing that such methods can hinder SSL model performance on cell-type prediction tasks. This effect grows more severe as the degree of data sparsity increases, underscoring the need for careful consideration of data handling strategies when deploying SSL models in practice.

An exciting revelation from the study was the significant enhancement in accuracy achieved by incorporating zero-shot random mask embeddings into advanced spatial clustering methodologies. This innovative integration suggests a promising pathway for improving the robustness of spatial clustering results, offering researchers new tools to refine their analyses in this emerging field. As spatial transcriptomics continues to evolve, the potential for SSL models to facilitate deeper insights into tissue architecture and cellular interactions stands as a beacon of possibility.

The implications of these findings extend beyond mere academic interest; they offer practical guidance for researchers striving to leverage pretrained models in their analyses of spatial transcriptomics data. By revealing both the capabilities and limitations of SSL models in cross-domain applications, the study serves as a roadmap for future investigations. The exploration of self-supervised learning within this context is not only timely but crucial, as it empowers scientists to optimize their methodologies and ultimately enhance our understanding of complex biological systems.

While SSL models have already shown impressive capabilities, the transferability of these models between distinct domains like scRNA-seq and spatial transcriptomics poses challenges that merit further investigation. As the community moves forward, addressing the intricacies of how these models can be adapted will be essential in promoting more accurate and nuanced analyses of biological data. Understanding how SSL can be effectively utilized across different data modalities is imperative to advancing our capacity to decode the intricacies of cellular environments.

As researchers continue to refine their approaches to model training and application, the timing is ripe for exploring the boundaries of what SSL can achieve. The robust performance of the random mask strategy SSL model is an encouraging indication that innovative methodologies are within reach. However, the juxtaposition of performance between models trained from scratch versus fine-tuned models obliges a deeper dive into the data characteristics that inform these outcomes, fostering a research environment where collaboration and curiosity thrive.

In conclusion, the study by Han et al. paves the way for future research aimed at elucidating the dynamics of self-supervised learning and its applicability in spatial transcriptomics. The interplay between model performance, data sparsity, and transferability may very well dictate the future trajectory of analytical methodologies in genomics research. With ongoing exploration into these questions, the potential for discovering new insights into cellular function and tissue organization remains vast and largely unexplored, urging the scientific community to push the boundaries of our understanding while harnessing the power of cutting-edge technologies.

In summary, as scientists continue to navigate the complexities of single-cell and spatial transcriptomics data, the resilience of SSL methodologies presents a promising frontier for innovation and discovery in biological research. The optimization of these models, understanding their limitations, and refining data handling techniques will undoubtedly forge new paths toward better comprehension of cellular behaviors and interactions, ultimately enriching our understanding of life at the molecular level.

With continued investigation and a commitment to bridging the gaps in data modalities, it is likely that self-supervised learning will unlock new doors in the study of gene expression and cellular heterogeneity, enriching our knowledge and capabilities in understanding the intricacies of biological systems and the frameworks that govern them.

Self-supervised learning has the potential to revolutionize our approach to data analysis in various fields, and the ongoing exploration of its applicability in spatial transcriptomics signifies a vital step in harnessing its full potential. As researchers persevere in overcoming the challenges associated with cross-domain model transferability, the insights gleaned from such studies will not only inform best practices but also inspire novel approaches that deepen our understanding of biology.

Adapting SSL methodologies for the unique challenges posed by spatial transcriptomics will not only yield immediate benefits but will also cultivate a rich environment for future innovations, drawing a closer connection between the analysis of cellular data and broader implications for health and disease. The excitement surrounding these developments is palpable, as the promise of such research endeavors holds the possibility of unveiling transformative insights into the workings of life itself.

Subject of Research: Transferability of self-supervised learning models from single-cell genomics to spatial transcriptomics.

Article Title: Reusability report: Exploring the transferability of self-supervised learning models from single-cell to spatial transcriptomics.

Article References:

Han, C., Lin, S., Wang, Z. et al. Reusability report: Exploring the transferability of self-supervised learning models from single-cell to spatial transcriptomics.
Nat Mach Intell 7, 1414–1428 (2025). https://doi.org/10.1038/s42256-025-01097-5

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s42256-025-01097-5

Keywords: self-supervised learning, spatial transcriptomics, single-cell RNA sequencing, model transferability, cell-type prediction, gene imputation, data sparsity, clustering methods.

Leveraging combined single-nucleus RNA sequencing (snRNA-seq), single-cell whole-genome sequencing (scWGS), and spatial transcriptomics, researchers meticulously mapped both genome-wide mutations and the corresponding transcriptomes of individual brain cells. This multi-faceted approach revealed a striking trend: short, highly expressed housekeeping genes—genes essential for core cellular functions—accumulate significantly more somatic single-nucleotide variants (sSNVs) over time. Intriguingly, this rise in mutation burden correlates strongly with a decrease in the expression levels of these crucial housekeeping genes.

A closer examination offers compelling evidence supporting this novel insight. Firstly, the enriched gene ontology terms related to housekeeping functions predominated among downregulated genes, particularly in neurons, whereas neuron-specific genes maintained relatively stable expression profiles during ageing. This suggests that ageing selectively impacts fundamental cellular maintenance processes rather than cell identity programs. Secondly, the study confirmed that housekeeping genes tend to be both short and robustly expressed, aligning with known genomic properties. Notably, the highest sSNV rates appeared in the shortest, most actively transcribed housekeeping genes.

Further statistical analysis unveiled a nuanced relationship capturing how these variables intersect. A multiple linear regression model demonstrated that elevated gene expression increased the likelihood of transcriptional downregulation with age, whereas longer gene length was associated with either preservation or even upregulation of transcriptional activity during ageing. These findings are particularly illuminating given the longstanding but inconsistent observations about gene length effects in ageing across various tissues. Within neurons, it appears that the transcriptional landscape favors retention of long, identity-defining genes while allowing somatic mutagenesis to erode short housekeeping genes.

The biological mechanisms driving these patterns are multifaceted. One plausible explanation posits that somatic mutations introduce premature stop codons or disrupt splicing fidelity, triggering nonsense-mediated decay pathways that reduce transcript abundance of affected genes. Additionally, faulty or aberrant DNA repair mechanisms implicated in the formation of somatic mutations could result in local epigenetic dysregulation, further influencing gene expression changes during ageing. Another fascinating aspect is the potential differential efficacy of DNA repair machinery between gene classes; short, highly expressed housekeeping genes may bear a higher burden due to their preferential engagement in transcription-coupled DNA repair (TCR).

Recent studies have revealed that single-stranded DNA lesions, often a precursor to mutations, can persist in human cells for extended durations without active repair, raising the possibility that transcription processes themselves may convert DNA damage into fixed, double-stranded mutations. Given that neurons are post-mitotic—non-dividing—and express high levels of topoisomerases that safeguard long genes, this cellular context may amplify the accumulation of mutations selectively in short housekeeping genes rather than the long neuron-specific genes.

Beyond these molecular insights, the research offers a rich portrait of cellular composition changes across the human lifespan. In infant brains, distinct populations of immature neurons and astrocytes were detected, along with an elevated ratio of oligodendrocyte precursor cells relative to their mature counterparts, supporting ongoing postnatal brain development. These developmental insights are complemented by the genomic profiling of somatic mutations in ageing neurons, which captured an increase in sSNVs with mutational spectra reminiscent of COSMIC mutational signatures SBS5 and SBS30, both previously linked to age-related mutagenesis and DNA damage responses.

Delving deeper, two novel mutational signatures designated A1 and A2 emerged from de novo analyses. Signature A1, characterized predominantly by T>C transitions, clustered with the clock-like SBS5 signature and showed enrichment in highly expressed genes, coding regions, and genomic loci marked by open chromatin. In contrast, Signature A2, dominated by C>T transitions and enriched in C>A and T>C variants associated with oxidative DNA damage, resembled the SBS30 signature but exhibited distinct enrichment in non-coding, repressed chromatin domains with repressive epigenetic marks.

The dynamic expression profiles of DNA base excision repair proteins, particularly NTHL1 and OGG1, within neurons across ageing provide tantalizing clues linking cellular repair activity to mutational signatures. While NTHL1’s decreased activity has been connected to SBS30 in other contexts, OGG1 involvement aligns with neuronal C>A mutations. The interplay between these repair pathways and accumulating somatic mutations likely shapes both the mutational landscape and transcriptomic alterations observed during neuronal ageing.

Crucially, the study’s comprehensive single-cell approach sets a new standard for exploring how somatic mutations intertwine with gene expression heterogeneity in distinct brain cell types. As single-cell whole-genome sequencing technologies continue to mature and expand across diverse cell populations, future investigations promise to unravel more intricate connections between somatic genomic alterations and functional consequences in the ageing brain.

This pioneering work not only deepens understanding of fundamental ageing biology but also paves the way for targeted interventions. By revealing vulnerabilities in housekeeping genes stemming from mutation accumulation and transcriptional changes, it opens potential therapeutic avenues aimed at preserving cellular homeostasis and delaying neurodegeneration. Moreover, the differential resilience of neuron identity genes hints at innate protective mechanisms that could be harnessed or augmented.

Altogether, this research exemplifies an integrative multi-omic leap forward in deciphering the genomic and transcriptomic choreography unfolding across human brain lifespan. It paints a detailed molecular narrative where mutation-driven erosion of essential housekeeping genes contrasts with preservation of cell identity programs, offering a refined lens through which to view ageing’s impact on brain health.

As the intersection of genetics, epigenetics, and transcriptomics continues to be illuminated at single-cell resolution, our grasp of brain ageing mechanisms will sharpen, enabling precision medicine strategies attuned to the unique vulnerabilities and strengths of neural circuits. The implication of transcription-coupled repair and mutational signatures further links genome maintenance processes to functional ageing, suggesting new biomarkers and targets for intervention.

Ultimately, this transformative research advances the frontier of neuroscience, calling attention to the silent genomic shifts that accumulate imperceptibly but inexorably within our brain cells, shaping cognition, resilience, and healthspan. Understanding these molecular changes is vital as populations age worldwide and the burden of neurodegenerative diseases rises, highlighting the promise of genomic and transcriptomic studies in the quest for healthier brain ageing.

Subject of Research: Single-cell transcriptomic and genomic changes during human brain ageing.

Article Title: Single-cell transcriptomic and genomic changes in the ageing human brain.

Article References:
Jeffries, A.M., Yu, T., Ziegenfuss, J.S. et al. Single-cell transcriptomic and genomic changes in the ageing human brain. Nature (2025). https://doi.org/10.1038/s41586-025-09435-8

Image Credits: AI Generated

The cornerstone of this research lies in the integration of single-cell RNA sequencing, single-cell multiomics, and spatial transcriptomics—cutting-edge methodologies that collectively map the complex cellular composition, gene expression profiles, and spatial organization within the prostate tumor microenvironment. These approaches enable a resolution previously unattainable in cancer biology, capturing the intricate interplay between diverse cell populations and their dynamic responses to therapeutic intervention. The study particularly focuses on the mechanisms that drive resistance to androgen deprivation therapy (ADT), the frontline treatment for advanced prostate cancer, which unfortunately succumbs to resistance in many patients.

Traditional models, including genetically engineered mice, have provided valuable insights into prostate cancer biology but fall short of representing the full spectrum of human disease progression, especially in the context of therapeutic resistance. Addressing this gap, the researchers employed these advanced single-cell techniques on mouse prostate tissues to dissect cellular heterogeneity and pinpoint the cell types responsible for tumor maintenance and adaptation following castration-mimicking androgen suppression. This comprehensive cellular cartography illuminates how distinct cell populations contribute to the tumor’s resilience and evolution under therapeutic stress.

One of the landmark findings from this research is the identification of over twenty genes whose activity is modulated in response to androgen deprivation. Notably, genes from the AP-1 and Klf families were significantly upregulated, revealing pathways likely involved in cellular stress response and the initiation of regenerative programs within the prostate tissue. Intriguingly, these gene expression patterns were mirrored in human prostate cancer samples from patients exhibiting resistance to androgen deprivation, underscoring the translational relevance of the murine model and the robustness of the cellular atlas produced.

The research team’s multiomic approach also uncovers how androgen deprivation therapy remodeling impacts the cellular ecosystem, reshaping intercellular interactions and signaling networks. This reconfiguration includes the activation of pathways associated with stress management and novel cell development, processes that potentially facilitate tumor cell survival amid a therapeutic assault. Such insights broaden our understanding of prostate cancer’s adaptive strategies and highlight potential vulnerabilities for future targeting.

Furthermore, the spatial transcriptomics data illuminate the precise anatomical contexts of these molecular changes within the prostate. By mapping where specific cell types and gene expression signatures localize, the study paints a vivid picture of tumor architecture and microenvironmental influences. This spatial dimension is crucial for identifying the niches that harbor resistant cancer cells and for designing localized therapeutic interventions that could disrupt these protective environments.

While many protein targets identified through this atlas are traditionally deemed difficult to drug due to their biological roles and molecular characteristics, the research team is actively exploring novel modalities to intervene in these pathways. These include designing molecules that can modulate protein-protein interactions, allosteric inhibitors, or emerging therapeutic platforms such as targeted protein degradation. This forward-looking strategy exemplifies how deep molecular understanding can guide innovative drug development in challenging cancer contexts.

The implications of this study extend beyond the scope of prostate cancer treatment resistance. It establishes a versatile framework for dissecting cellular ecosystems in cancer and other diseases, emphasizing the power of integrating multiomic data with spatial context. This comprehensive approach sets a precedent for future research endeavors seeking to unravel the complexity of tumor biology and therapeutic response at an unprecedented resolution.

The lead investigators emphasize that their work not only reveals the hidden diversity within prostate cell populations but also exposes the cellular programs that empower tumor survival against one of the most effective current therapies. By providing a detailed roadmap of resistance mechanisms, this research opens avenues for the rational design of next-generation treatments aimed at preventing or overcoming castration resistance—a clinical hurdle that has limited the efficacy of androgen deprivation therapy for decades.

Looking ahead, the team plans to extend their cellular atlas to human prostate tissue samples. This next phase promises to refine the catalog of biomarkers indicative of treatment response and resistance, potentially enabling personalized therapeutic strategies tailored to the molecular landscape of individual tumors. Such advancements could revolutionize the clinical management of prostate cancer, shifting from reactive to proactive, precision-guided treatment approaches.

In sum, this integrative study leverages cutting-edge technologies to unravel the cellular and molecular fabric of prostate cancer progression under androgen deprivation therapy. The findings underscore the complexity of tumor adaptation and provide a rich repository of targets for future therapeutic exploration. By illuminating the pathways that confer treatment resistance, this work heralds a new era in prostate cancer research and therapy development, holding promise to improve prognosis and quality of life for countless patients worldwide.

Subject of Research: Cells

Article Title: Cellular cartography reveals mouse prostate organization and determinants of castration resistance

Web References:
https://doi.org/10.1073/pnas.2427116122

References:
“Cellular cartography reveals mouse prostate organization and determinants of castration resistance,” Proceedings of the National Academy of Sciences, DOI: 10.1073/pnas.2427116122

Image Credits:
Jacob Dwyer, Justine Ross, Michigan Medicine

Keywords:
Health and medicine