The study, published in Frontiers in Cell and Developmental Biology, addresses a persistent gap in gynecological oncology. Endometrial carcinoma is the most common gynecological malignancy in high-income countries, and roughly eighty percent of patients are diagnosed with FIGO stage I–II disease. About eighty-five percent of these tumors are the endometrioid subtype. Current risk assessment relies on stage, histological grade, depth of myometrial invasion, lymphovascular space invasion, and molecular subtype assignment into POLE-mutant, microsatellite-instability, copy-number-low, and copy-number-high classes. Even with molecular classification, which has measurably improved risk assignment, clinically relevant uncertainty remains. Histological grading suffers from interobserver variability, lymphovascular space invasion is difficult to distinguish from morphological mimics, and molecular subtypes must still be interpreted alongside tumor architecture, spatial heterogeneity, and clinical context.

The new framework, formally named the prototype-resolved interpretable survival model, was built to integrate routine haematoxylin and eosin-stained whole-slide images with RNA-sequencing-derived pathway programmes for progression-free survival prediction. The technical design is distinctive. Diagnostic slides are cut into 256-by-256-pixel patches at an equivalent twentyfold magnification, and a pathology encoder converts each patch into a numerical embedding, producing five hundred image tokens per patient. On the molecular side, gene expression is filtered, log-transformed, standardized, and organized into 236 pathway-level tokens drawn from Hallmark and KEGG gene sets. The critical innovation is an adaptive prototype-learning bottleneck that compresses the hundreds of image tokens into just four learnable morphological prototypes — shared tissue states that recur across patients — before any cross-modal fusion occurs.

This compression matters because directly modelling the interactions between hundreds of image patches and hundreds of molecular pathways creates an enormous parameter space that is difficult to estimate reliably in disease-specific cohorts with limited matched data. By first distilling tissue appearance into four prototypes and then aligning those prototypes with pathway-level representations, PRISM reduces the multimodal fusion problem to a manageable size while retaining traceability back to specific regions of the slide. A pathway–prototype alignment module links transcriptional programmes to morphological states, and the fused representation feeds a Cox proportional-hazards output head that produces a continuous risk score, with higher scores indicating greater probability of progression, recurrence, or death.

The development process was deliberately rigorous. The model was trained on 229 patients from The Cancer Genome Atlas uterine carcinoma cohort, with a held-out internal validation set of 58 patients. Only after the architecture, hyperparameters, preprocessing workflow, and checkpoint-selection rule were completely locked was the model retrained on all 287 TCGA cases and evaluated exactly once, without any tuning, in an independent external cohort of 63 patients from the Clinical Proteomic Tumor Analysis Consortium. The CPTAC data played no role whatsoever in training, feature selection, or threshold optimization — a design feature that lends unusual credibility to the external results in a field where leakage between discovery and validation sets is a common criticism.

Across 350 total patients, PRISM achieved Harrell’s concordance indices of 0.770 in the TCGA training cohort, 0.751 in internal validation, and 0.745 in the external CPTAC cohort, indicating that risk ranking held up when transported to a different patient population, sequencing platform, and slide-scanning environment. At the thirty-six-month horizon, the censoring-adjusted time-dependent area under the curve reached 0.727 externally, compared with 0.685 for a baseline model combining clinicopathological variables and molecular subtype. In a complementary fixed-horizon analysis restricted to patients whose thirty-six-month event status was ascertainable, adding PRISM to the clinicopathological model raised the AUC in the external cohort from 0.609 to 0.719. In the TCGA evaluable subset, the corresponding jump was even larger, from 0.697 to 0.905 when molecular subtype was included. Ablation experiments reinforced the architecture’s logic: a version without the prototype bottleneck and an image-only model both performed worse and varied more across five independently generated data splits, while the image-only model collapsed to a concordance index of 0.584 in external validation.

Perhaps the most striking feature of the study is its interpretability. Because the model’s risk predictions pass through named morphological prototypes, a gynecological pathologist could review the patches assigned to each prototype and attach descriptive labels. Prototype one corresponds predominantly to stromal and fibromuscular tissue, prototype two to tumour epithelial and glandular morphology, and prototype three to hypercellular, pleomorphic tumour regions. A fourth prototype captured low-content and artefact-associated tissue edges and was excluded from biological interpretation. Spatial activation maps showed that the first three prototypes occupied distinct regions within individual slides, meaning the model’s risk estimate could be traced to recognizable tissue compartments. High-risk tumors showed lower stromal prototype burden and higher hypercellular prototype burden — a shift from supportive tissue architecture toward aggressive cellular morphology.

The molecular side of the interpretation proved equally coherent. Tumours classified as high risk by PRISM showed elevated pathway-programme scores for proliferation, cell-cycle activity, glycolysis, hypoxia, focal adhesion, and epithelial–mesenchymal transition, while low-risk tumors scored higher on metabolism- and homeostasis-associated programmes. These are precisely the biological themes that independent studies of endometrial cancer have implicated in aggressive behavior. The pathway–prototype alignment patterns were strongly concordant across the two cohorts, with a pooled Pearson correlation of 0.91, and within-prototype correlations ranged from 0.44 to 0.73. When the TCGA-derived risk cut-point was applied unchanged to the CPTAC cohort, Kaplan–Meier curves separated significantly, and the continuous risk score remained associated with progression-free survival after multivariable adjustment for age, grade, molecular subtype, invasion depth, radiation therapy, and residual tumour status.

The authors are candid about limitations. The external cohort contained only 63 patients and eight progression events, so the confidence intervals around external discrimination gains span zero, and the external calibration analysis showed attenuation and overprediction of absolute risk, making recalibration a necessary step before any clinical transport. The risk cut-point was selected and evaluated within the same TCGA cohort, rendering that analysis descriptive, and each patient was represented by a single diagnostic slide. The multivariable external association is explicitly framed as exploratory rather than definitive evidence of independent prognostic value. The model also requires transcriptomic input, which is not universally available at diagnosis, and the pathway scores and prototype assignments remain model-derived constructs that will need confirmation through spatial transcriptomics, single-cell sequencing, and functional studies.

Nevertheless, the work represents a meaningful proof of concept for prototype-guided morphomolecular risk modelling in a clinically relevant early-stage population. The locked external evaluation, repeated-split sensitivity analysis, and spatially traceable outputs provide a well-defined foundation for the larger, prospective, multi-institutional validation the authors say is warranted before deployment. More broadly, the study raises a question with implications well beyond endometrial cancer: which morphomolecular relationships — the recurring partnerships between how tissue looks and what its cells are doing transcriptionally — remain stable across institutions and tumor types, and which vary with lineage and context? If recurrent adverse programmes such as proliferation, hypoxia, and epithelial–mesenchymal transition are expressed through shared tissue states across cancers, a comparative atlas of prototype–pathway maps could improve model transportability everywhere and clarify, at last, how molecular risk is written into the architecture of tissue itself.

Subject of Research: Interpretable multimodal AI integrating histopathology and transcriptomics for progression-risk stratification in early-stage endometrioid endometrial carcinoma

Article Title: An interpretable multimodal morphomolecular model refines progression-risk stratification in early-stage endometrioid endometrial carcinoma

Article References: Li, J., Tang, Q., Chen, J., Chen, Q., Liu, B., Kong, Q., Ji, J., & Zhao, Q. (2026). An interpretable multimodal morphomolecular model refines progression-risk stratification in early-stage endometrioid endometrial carcinoma. Frontiers in Cell and Developmental Biology, 14, Article 1945984. https://doi.org/10.3389/fcell.2026.1945984

Image Credits: AI Generated

DOI: 10.3389/fcell.2026.1945984

Keywords: endometrial cancer, computational pathology, multimodal AI, risk stratification, transcriptomics, whole-slide imaging, prototype learning, progression-free survival, molecular subtyping, interpretable machine learning, TCGA, CPTAC

The motivation behind the study lies in a practical bottleneck of modern oncology. Molecular profiling, which guides targeted therapies and immunotherapy decisions, currently depends on genomic or transcriptomic assays such as next generation sequencing. These tests are expensive, require specialized infrastructure, and can add days or weeks to the diagnostic timeline, a delay that matters enormously for patients with aggressive disease. If a machine could reliably estimate molecular features from the slide that is already being examined for diagnosis, the turnaround time for biomarker information could shrink dramatically and the cost per test could fall to nearly nothing beyond the computational expense. That is the promise the authors set out to test on a pan cancer scale rather than in a single tumor type.

The system, whose name stands for Rapid and Intelligent Detector for Genetic Estimation, is built on weakly supervised learning, a strategy that sidesteps one of the most stubborn obstacles in computational pathology. Whole slide images are gigantic, often exceeding one hundred thousand by one hundred thousand pixels, and labeling specific regions of interest by hand is prohibitively laborious. RIDGE instead learns from slide level labels only, using a multiple instance learning formulation in which the slide is treated as a bag of smaller tissue patches and only the overall slide carries a molecular label. The architecture employs fully convolutional networks to aggregate patch level information, allowing the model to identify which morphological patterns are associated with a given biomarker without ever being told where to look.

Several technical choices distinguish the framework. The authors incorporate clustering guided contrastive learning, a self supervised pretraining approach that teaches the network to group visually and biologically similar tissue patterns together before any biomarker prediction begins. Attention mechanisms, including a focused linear attention module, allow the model to weigh the contribution of thousands of patches efficiently, while depthwise convolutions and a mixture of experts design help the network specialize across the heterogeneous landscape of tumor types. Multi task learning enables a single model to predict multiple biomarkers and gene expression signatures simultaneously, sharing learned representations across related prediction problems. The result is a general purpose system intended to work across cancers rather than a bespoke model for each disease.

The training data were substantial. The team developed and validated RIDGE using 4,983 whole slide images from 4,680 patients spanning 12 solid tumor types in The Cancer Genome Atlas, including breast, lung, colon, rectal, stomach, liver, pancreatic, cervical, head and neck, and three kidney cancer cohorts. Performance was measured with the area under the receiver operating characteristic curve, the standard metric for binary classification tasks in medicine. Across all 12 cancer types, RIDGE achieved an overall AUC of 0.763, with a 95 percent confidence interval of 0.724 to 0.802. In a field where individual biomarker prediction models often hover in a similar range, a single framework reaching this level across such a diverse set of tumors and molecular targets is a meaningful benchmark.

Perhaps the most important result concerns generalization beyond the training data. Machine learning models in medicine frequently fail when moved to new datasets, a phenomenon driven by differences in staining protocols, scanners, and patient populations. To test reproducibility, the researchers applied RIDGE to an external validation cohort of 221 whole slide images from 105 colorectal cancer patients in the Clinical Proteomic Tumor Analysis Consortium, a completely independent resource, asking the model to predict microsatellite instability status. MSI is a critical biomarker because microsatellite unstable tumors respond well to immune checkpoint inhibitors. RIDGE achieved an AUC of 0.769, with a confidence interval of 0.686 to 0.841, closely matching its internal performance and suggesting the learned features reflect genuine biology rather than dataset specific artifacts.

Beyond raw accuracy, the authors emphasize interpretability, a persistent concern for clinicians asked to trust black box algorithms. The study reports that the model captures morphological visual characteristics that make gene expression signatures detectable directly from the slides, and supplementary analyses include explainability experiments examining how molecular features manifest in gastric cancer tissue. In effect, the network learns to associate particular cellular and tissue architectures, such as the appearance of tumor infiltrating immune cells, gland formation patterns, or nuclear features, with underlying molecular states. This capacity to quantify genotype phenotype relationships from images could itself become a research tool, allowing investigators to map molecular biology onto tissue morphology at a scale that manual review could never achieve.

The clinical implications, if the approach matures, are considerable. Because H&E stained slides are universally produced, an AI layer on top of standard pathology workflows could provide preliminary biomarker estimates within minutes of slide scanning, flagging patients who should receive priority for confirmatory molecular testing. In hospitals without access to sequencing facilities, such predictions could guide referral decisions and broaden equitable access to precision oncology. The authors argue that RIDGE could significantly expedite cancer screening and personalized therapy, and the pan cancer design means a single deployment could serve pathology departments handling many tumor types rather than requiring separate pipelines for each indication.

Caution is nonetheless warranted before such systems reach the clinic. An AUC of roughly 0.76, while respectable, indicates imperfect discrimination, meaning the model would need to function as a triage and prioritization tool rather than a replacement for definitive molecular assays. The study relied on publicly available, de identified data from TCGA and CPTAC, which, although multi institutional, may not capture the full heterogeneity of staining practices and patient demographics seen in routine global practice. Prospective clinical validation, regulatory review, and demonstration of impact on patient outcomes remain necessary steps. The published work was exempt from additional ethics approval because it used existing consented datasets, but real world deployment would face a new and more demanding evaluation landscape.

Even with those caveats, the study adds to a rapidly growing body of evidence that the humble pathology slide is an information rich object whose molecular content can be unlocked computationally. By demonstrating a single weakly supervised framework that predicts multiple biomarkers across a dozen cancer types and reproduces its performance on external data, the RIDGE team has moved the field closer to a future in which every diagnostic slide yields both a morphological diagnosis and a molecular profile. For patients, that could mean faster answers at lower cost; for researchers, a powerful new lens on the relationship between how a tumor looks and what it is, at the level of its genes.

Subject of Research: Deep learning prediction of pan-cancer molecular biomarkers from H&E-stained whole slide pathology images

Article Title: Deep learning-based large-scale pan-cancer multiple biomarkers prediction using RIDGE with pathological images

Article References: Xi, H., Feng, X., Lu, Y., Li, G., Zhang, Y., Li, J., Wang, Y., Xu, J., Zhang, Y., Sha, C., & He, M. (2026). Deep learning-based large-scale pan-cancer multiple biomarkers prediction using RIDGE with pathological images. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02695-4

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02695-4

Keywords: computational pathology, deep learning, whole slide imaging, biomarker prediction, weakly supervised learning, multiple instance learning, microsatellite instability, TCGA, CPTAC, precision oncology, gene expression, pan-cancer

The team’s central question was deceptively simple: does the N-glycosylation of extracellular matrix proteins, the enzymatic attachment of complex glycans to asparagine residues, provide information about cancer-associated fibroblast biology that goes beyond simply knowing how much of each protein is present? To answer it, the investigators turned to one of the richest public resources in cancer proteomics, the Clinical Proteomic Tumor Analysis Consortium pancreatic cancer dataset. They integrated peptide-level and site-level N-glycoproteomics from 135 pancreatic ductal adenocarcinoma tumors with matched proteomics, RNA expression data, clinical covariates, external transcriptomic cohorts, and spatial transcriptomics. Five adenosquamous tumors were retained only for sensitivity analyses, ensuring that the main findings reflected the conventional ductal histology.

The methodological heart of the study lies in a statistical maneuver called protein-residualization. For each ECM glycopeptide and glycosylation site, the researchers regressed out the abundance of the matched parent protein, leaving behind what they term protein-residual ECM N-glycosylation scores. In plain terms, they asked whether a given ECM protein is more or less heavily glycosylated than its own abundance would predict, and whether that excess or deficit of glycosylation tells a coherent biological story. This is a post-translational readout: it captures enzymatic and cellular regulation of glycan assembly that cannot be inferred from gene expression or raw protein levels alone. The approach is conceptually analogous to looking not at how many bricks were delivered to a construction site, but at how those bricks were finished and treated, which may reveal which crew is doing the building.

What emerged was a striking and reproducible association. Residual ECM N-glycoprotein scores were significantly linked to the conventional myofibroblastic cancer-associated fibroblast phenotype, known as myCAF, with adjusted beta coefficients of 0.426 and 0.425, and to broader CAF and stromal phenotypes with betas of 0.304 and 0.312. Critically, these associations survived adjustment for the proteome matrisome score, stromal fraction, and neoplastic cellularity, meaning the glycosylation signal was not simply a disguised measure of having more stroma or fewer tumor cells. The signal was distributed across a candidate program of ECM glycoproteins rather than driven by a single molecule, which strengthens the interpretation that it reflects a coordinated stromal cell state rather than a technical artifact.

Robustness was tested with unusual thoroughness. The authors evaluated nonlinear residualization schemes, detection-probability weighting, feature coverage thresholds, mass spectrometry acquisition plex, and random feature resampling. The myCAF association remained stable under nonlinear residualization, detection-probability weighting, observation thresholds ranging from 40 to 110 tumors, and random splitting of the glycosylation features. In an era when many multi-omics associations dissolve under scrutiny, this level of stress-testing lends considerable weight to the central claim. The team also examined CAF subtypes, glycan composition, spatial localization, and clinical context, building a multi-angled case that the glycosylation signature is a genuine biological phenomenon.

Among the composition-resolved findings, one stands out. When glycoforms were classified by glycan composition, high-mannose glycoforms on ECM proteins showed the strongest myCAF effects, with beta coefficients of 0.552 and 0.549. High-mannose N-glycans are relatively immature, unprocessed structures that can accumulate when glycan processing enzymes are altered or when proteins traffic through the secretory pathway differently. Their enrichment in the myCAF-associated stromal state suggests that myofibroblastic fibroblasts in pancreatic tumors may impose a distinctive glycan-processing environment on the matrix they deposit, and it nominates these glycoforms as concrete, testable translational candidates.

Spatial evidence provided independent confirmation. Using GeoMx digital spatial profiling, the researchers compared CAF-rich segments against epithelial segments in 20 patients and found that the candidate RNA program corresponding to the glycosylation signature was higher in the CAF segments in every single patient, with a mean paired difference of 1.234, a 95 percent bootstrap confidence interval of 1.054 to 1.389, and a p value of 9.57 times 10 to the minus 5. The corresponding candidate genes also localized to CAF-rich spatial niches in spatial transcriptomic data, and the pattern held across external transcriptomic cohorts, including resources from The Cancer Genome Atlas and the Gene Expression Omnibus. Convergence of proteomic, transcriptomic, and spatial lines of evidence on the same fibroblast state is exactly the kind of triangulation that translational scientists look for before investing in experimental validation.

The biological logic of the finding is worth unpacking. Cancer-associated fibroblasts in pancreatic cancer exist in several recognized states, including inflammatory iCAFs, antigen-presenting apCAFs, and contractile, matrix-producing myCAFs. MyCAFs are closely apposed to invasive tumor cells, deposit dense collagenous matrix, and have been linked in many studies to tumor progression, immune exclusion, and treatment resistance. If the glycosylation state of the ECM they build is a faithful readout of their activity, then measuring ECM glycopeptides offers a way to fingerprint the myCAF compartment in bulk tumor samples, something that single-cell and spatial methods capture but that has been difficult to access in standard proteomic workflows. In effect, the tumor’s sugar-coated scaffolding becomes a recording device for which fibroblast program dominated its construction.

The translational implications are twofold. First, protein-residual ECM N-glycosylation scores could serve as biomarkers that stratify pancreatic cancer patients by stromal biology, potentially identifying tumors dominated by myCAF activity without requiring spatial assays on every specimen. Second, the specific nomination of high-mannose ECM glycoforms and recurrent ECM glycoproteins gives glycosite-resolved experimental validation a concrete starting point. The authors are careful to frame these as candidates for validation rather than validated targets, and the study’s design, being based on public consortium data, means that functional experiments in model systems will be the necessary next step. Still, the fact that the signal survived such extensive robustness analyses and replicated across independent data modalities makes it a credible foundation for that work.

More broadly, the study contributes to a growing recognition that glycosylation is not biochemical decoration but a functional dimension of the tumor microenvironment. Glycans influence protein folding, stability, interactions with lectins, and immune recognition, and the ECM is a glycan-rich landscape that tumors remodel extensively. By showing that a post-translational, composition-resolved layer of the matrix encodes fibroblast state information invisible to protein abundance measurements, the researchers open a path toward glycoproteomics-informed stromal profiling in other fibrotic and desmoplastic tumors as well. For a disease like pancreatic cancer, where the stroma has long been both an obstacle and an enigma, the demonstration that its sugar signatures can be read, quantified, and mapped to a specific cellular state is a small but potentially consequential step toward therapies that target the ground the tumor stands on rather than the tumor alone.

Subject of Research: Protein-residual ECM N-glycosylation as a readout of myCAF-enriched stromal states in pancreatic ductal adenocarcinoma

Article Title: Protein-residual ECM N-glycosylation delineates a composition-resolved myCAF-enriched stromal state in pancreatic ductal adenocarcinoma

Article References: Feng, Y., Li, Z., Wang, H., Peng, Z., Tang, Y., Wei, Q., & Liu, H. (2026). Protein-residual ECM N-glycosylation delineates a composition-resolved myCAF-enriched stromal state in pancreatic ductal adenocarcinoma. Journal of Translational Medicine. https://doi.org/10.1186/s12967-026-08974-6

Image Credits: AI Generated

DOI: 10.1186/s12967-026-08974-6

Keywords: pancreatic ductal adenocarcinoma, N-glycosylation, glycoproteomics, extracellular matrix, cancer-associated fibroblasts, myCAF, high-mannose glycoforms, spatial transcriptomics, CPTAC, tumor microenvironment, biomarkers, Journal of Translational Medicine