The study’s methodological core is as notable as its biological findings. The researchers began with single-cell RNA sequencing datasets from patients with non-small cell lung cancer who had undergone immunotherapy, drawing on three publicly available cohorts catalogued in the Gene Expression Omnibus under the accession numbers GSE207422, GSE179994, and GSE176021. After careful annotation against a large-scale pancancer atlas and established hallmarks of cancer biology, they resolved the tumor microenvironment into 109 distinct cell types, a granularity that captures not only malignant cells but also the intricate cast of immune and stromal players that surround them, from conventional dendritic cells and tumor-infiltrating myeloid cells to cancer-associated fibroblasts.

What sets the work apart is how the team interrogated this cellular landscape. Rather than relying solely on conventional differential-expression analysis, which can only report correlations, they deployed Geneformer, a transformer-based foundation model trained on large corpora of single-cell transcriptomic data. Geneformer allows what the authors call in-silico single-gene knockouts: the computational deletion of a gene across every cell type in the dataset, followed by prediction of how that deletion reshapes the cell’s transcriptional state. By systematically performing these virtual perturbations across all 109 cell types, the researchers could estimate each gene’s functional influence on immunotherapy outcome, a task that would be experimentally prohibitive if attempted gene by gene in the laboratory.

This virtual screen converged with more traditional approaches to yield a ranked list of 110 genes associated with immunotherapy efficacy. Of these, 25 were linked to better treatment outcomes, while 85 were associated with worse outcomes, a skew that reflects the biological reality that many tumor programs actively suppress immune attack. Crucially, the analysis did not treat the tumor as a uniform mass. The team mapped precisely which cell types each gene acted in, revealing that the same gene can exert opposing or complementary effects depending on whether it is expressed in malignant cells, cytotoxic lymphocytes, or stromal compartments. Hierarchical clustering then organized these genes into functional modules, and variance-partitioning statistics quantified how much of the variation in immunotherapy response between patients each gene could explain.

To translate this catalog into something clinically usable, the researchers built an immune scoring system based on all 110 genes, and then distilled it into a leaner model called CancerCellScore, built on just seven genes. GSTP1 emerged as the top contributor among those seven. The scoring system was verified to distinguish patients who respond to immunotherapy from those who do not, offering a potential decision-support tool for oncologists weighing whether checkpoint blockade is the right first move. To make the tool accessible, the team deployed a website at immunotherapy.live/ResisGenes, where the 110 genes can be visualized and real-time sample scores can be computed, an unusual degree of transparency and practicality for a discovery-stage biomarker study.

The mechanistic story that follows is where the study acquires its therapeutic punch. GSTP1 encodes glutathione S-transferase Pi 1, a detoxification enzyme best known for conjugating electrophilic compounds to glutathione and thereby shielding cells from oxidative damage. The researchers found that GSTP1 expression was negatively associated with immunotherapy response: the more of it a tumor expressed, the less likely the patient was to benefit from anti-PD-1 therapy. This connection is not arbitrary. GSTP1 sits squarely within the antioxidant machinery that determines whether a cancer cell succumbs to ferroptosis, an iron-dependent form of regulated cell death driven by the accumulation of lipid peroxides, a process in which the lipid peroxidation marker 4-hydroxynonenal serves as a classic readout.

By keeping glutathione levels high and lipid peroxidation in check, GSTP1 effectively buffers tumor cells against the oxidative stress that ferroptosis-inducing immune pressure generates. The team validated this logic in the laboratory using RNA interference approaches, including short hairpin RNAs and small interfering RNAs, and confirmed the regulatory circuitry with chromatin immunoprecipitation and quantitative polymerase chain reaction, tracing GSTP1’s activity to antioxidant response elements in its regulatory regions. The most striking result, however, came from in vivo experiments: when anti-GSTP1 strategies were combined with anti-PD-1 therapy, tumor growth was significantly inhibited compared with either approach alone. In other words, disarming GSTP1 appeared to re-sensitize tumors to checkpoint blockade, plausibly by lowering the ferroptosis threshold and allowing the immune-stimulated oxidative assault to push cancer cells over the edge into death.

The implications for patients with non-small cell lung cancer are twofold. First, the seven-gene CancerCellScore model, with GSTP1 at its head, could serve as a predictive biomarker, identifying before treatment begins which patients are likely to resist PD-1 inhibitors and might instead be candidates for combination regimens or clinical trials. Second, GSTP1 itself becomes a druggable target. Glutathione S-transferases have long interested pharmacologists because of their role in chemotherapy detoxification, and the present study reframes that pharmacology in the immunotherapy era: inhibiting GSTP1 is not merely a way to make tumors more chemosensitive, but a way to make them more visible and vulnerable to the immune system. The study also received ethical approval from the Sun Yat-sen University Cancer Center Animal Care and Use Committee, underscoring that the in vivo findings rested on formally reviewed animal work.

There are, of course, the usual caveats that separate a compelling translational study from bedside practice. The scoring system was developed and verified in retrospective cohorts, and prospective validation in independent patient populations will be needed before it can guide treatment decisions. The combination of GSTP1 targeting with anti-PD-1 therapy, while dramatically effective in preclinical models, must clear the familiar hurdles of pharmacology, safety, and tumor delivery before human trials can be designed with confidence. The authors themselves frame the work as a discovery and prioritization engine rather than a finished therapeutic protocol, and the open-access release of the model, the gene list, and the live scoring website invites the wider community to stress-test the findings against their own datasets.

Even with those caveats, the study stands as a vivid demonstration of how large language models adapted to biology are changing the pace and scale of target discovery. A transformer trained on single-cell data can, in silico, delete thousands of genes across more than a hundred cell types and rank their contributions to a clinical outcome, compressing years of bench work into a computational pass that then directs focused experiments. The result here is a complete arc, from machine-learning perturbation screens through multi-omic prioritization to mechanistic validation and a candidate combination therapy, all centered on a single enzyme that lung tumors use to shield themselves from both oxidative death and immune destruction. If GSTP1 inhibition fulfills its promise in the clinic, the path that led there, virtual knockouts across 109 cell types converging on ferroptosis as the Achilles heel of immunotherapy resistance, may prove as influential as the target itself.

Subject of Research: GSTP1-mediated resistance to anti-PD-1 immunotherapy in non-small cell lung cancer identified by transformer-based single-cell perturbation modeling

Article Title: GSTP1 is associated with NSCLC immunotherapy resistance: multi-omic discovery via transformer-based perturbation across 109 cell types

Article References: GSTP1 is associated with NSCLC immunotherapy resistance: multi-omic discovery via transformer-based perturbation across 109 cell types. (n.d.). https://doi.org/10.1186/s12967-026-08949-7

Image Credits: AI Generated

DOI: 10.1186/s12967-026-08949-7

Keywords: non-small cell lung cancer, immunotherapy resistance, GSTP1, single-cell RNA sequencing, Geneformer, in-silico knockout, anti-PD-1, ferroptosis, biomarker, tumor microenvironment, transformer model, Journal of Translational Medicine

The motivation behind the work lies in a paradox that has emerged from large-scale plasma proteomics studies. Over the past several years, researchers measuring thousands of proteins in blood samples from tens of thousands of individuals have shown that these molecular fingerprints can estimate chronological age with striking accuracy. Yet a protein that predicts age is not necessarily a protein that drives ageing. Correlation, as the field has repeatedly learned, does not equal causation, and the leap from a predictive biomarker to a druggable target requires layers of evidence that simple statistical associations cannot supply. PEN was designed specifically to make that leap more systematic, transparent, and reproducible.

Architecturally, PEN unfolds in six sequential stages, each of which adds a distinct form of biological reasoning. The pipeline begins with a plasma proteome input, typically a matrix of protein abundance measurements across many individuals. In the second stage, a multilayer perceptron, a classic type of deep neural network, is trained to predict chronological age from the protein measurements. In the demonstration study, the team applied this stage to data from 44,179 UK Biobank participants profiled for approximately 2,920 plasma proteins. The model achieved a test coefficient of determination of 0.8723, a Pearson correlation of 0.934, and a mean absolute error of just 2.30 years, confirming that the proteomic age clock was performing at a level comparable to the best published proteomic predictors.

The third stage is where PEN begins to move beyond prediction. Using a technique called Integrated Gradients, an attribution method originally developed for interpreting deep learning models, the researchers identified which proteins contributed most strongly to the age predictions and characterized the functional modules those proteins participate in. Integrated Gradients works by accumulating the gradients of the model’s output with respect to its inputs as features are gradually shifted from a baseline to the actual value, yielding a per-protein importance score. This step converts the opaque neural network into a ranked list of diagnostic proteins, providing the raw material for the target discovery stages that follow.

Stage four introduces the graph-based core of the framework. The diagnostic proteins are mapped onto a protein-protein interaction network, a vast molecular map in which nodes represent proteins and edges represent known physical or functional interactions. Graph convolutional propagation and diffusion algorithms then spread information across this network, allowing the framework to identify candidate targets that are not merely important on their own but are embedded in neighborhoods of ageing-relevant biology. This network context matters because biological systems rarely act through single molecules; perturbing one node of a densely connected module can ripple through entire pathways. By propagating evidence through the interactome, PEN surfaces candidates that a purely statistical analysis would likely overlook.

The fifth stage applies a discovery-oriented reranking that integrates multiple components of biological evidence, weighing each candidate against additional criteria before a final prioritization is issued. Applied to the UK Biobank plasma proteomics data, this process produced a shortlist of candidate mechanistic target hypotheses that includes FBN1, FBLN5, EDA, RLN3, LHCGR, COL14A1, HLA-E, and TSPAN4. Several of these names will be familiar to students of ageing biology. FBN1 encodes fibrillin-1, a structural component of elastic fibers, and FBLN5 encodes fibulin-5, another extracellular matrix protein essential for elastic fiber assembly, both of which connect plausibly to the vascular stiffening and tissue degeneration characteristic of ageing. HLA-E, an immune regulatory molecule, hints at the inflammatory and immunological dimensions of the ageing process.

The sixth and final stage provides what the authors describe as orthogonal in silico support through virtual perturbation. Here the framework turns to Geneformer, a foundation model of gene regulatory networks trained on large corpora of single-cell transcriptomic data. Geneformer can simulate, computationally, what happens to a cell’s transcriptional state when a particular gene’s activity is reduced or abolished, a process the researchers call virtual perturbation. The team applied this approach to an independent dataset, GSE130973, a publicly available single-cell RNA sequencing dataset of ageing human skin. By asking whether virtual perturbation of each candidate gene shifts cells from an OLD transcriptional state toward a YOUNG one, or vice versa, the framework obtained an independent line of computational evidence that is entirely separate from the plasma proteomics data used to generate the candidates.

The results of this final stage were instructive in both their successes and their limitations. After recalibrating the analysis using an expanded null distribution of 500 random token-valid genes, FBLN5 and HLA-E showed significant positive OLD-to-YOUNG transcriptional state shifts in specific cell populations, strengthening the case that these two proteins merit experimental attention. Other candidates, including several on the original shortlist, displayed weaker or context-dependent effects, a reminder that computational prioritization is a hypothesis-generating exercise rather than a guarantee of biological validity. The authors are explicit on this point: the Geneformer perturbation stage strengthens biological plausibility but does not constitute experimental validation. Laboratory work with cell models, organoids, or animal systems remains essential before any of these candidates can be considered genuine therapeutic targets.

What makes PEN notable is less any single algorithmic component than the discipline of its overall design. Each stage is modular and auditable, and the framework is explicitly intended to be reproducible and generalizable. The authors argue that the same six-stage logic, proteomic input, deep learning prediction, interpretability analysis, graph-based propagation, evidence-weighted reranking, and foundation-model perturbation testing, could be applied to other biomarker-driven target discovery problems far beyond ageing, from cardiometabolic disease to neurodegeneration. The study was conducted using the UK Biobank Resource under Application Number ID200882 and was supported by the Zhangjiang Special Funding Major Project 2024, an initiative titled AI-driven and multi-ancestry evaluated protein therapeutic discovery system, reflecting a broader industrial and academic push to convert population-scale omics data into actionable drug discovery pipelines.

The broader significance of the work lies in its timing. Proteomic age clocks are proliferating, foundation models for biology are maturing rapidly, and the pharmaceutical industry is increasingly willing to pursue ageing itself, or age-related frailty, as a therapeutic domain. What has been missing is a principled way to connect these pieces, to move from a blood test that knows how old you are to a molecule worth targeting in the clinic. PEN offers one template for that connection, complete with built-in humility about what computation can and cannot establish. If candidates such as FBLN5 and HLA-E survive the crucible of experimental validation, the framework that nominated them may well become a standard fixture in the computational geroscience toolkit, and the era in which a blood draw could point directly toward an anti-ageing intervention will have moved one decisive step closer.

Subject of Research: A graph-guided computational framework for prioritizing ageing-associated drug target hypotheses from plasma proteomics data.

Article Title: Protein Expression Net: an integrated graph-guided computational framework and virtual perturbation approach for prioritizing ageing-associated target hypotheses from plasma proteomics

Article References: Wu, W., Jiang, Y., Wei, K., Lin, Z., Gu, R., Wei, S., Wang, Z., Fang, L., Wang, X., Fu, X., Wang, Y., & Pan, L. (2026). Protein Expression Net: an integrated graph-guided computational framework and virtual perturbation approach for prioritizing ageing-associated target hypotheses from plasma proteomics. BMC Bioinformatics. https://doi.org/10.1186/s12859-026-06645-3

Image Credits: AI Generated

DOI: 10.1186/s12859-026-06645-3

Keywords: ageing, plasma proteomics, UK Biobank, deep learning, graph neural network, Integrated Gradients, Geneformer, virtual perturbation, target prioritization, BMC Bioinformatics, protein-protein interaction network, biomarkers

The research, led by Ruizhe Huang, Zhiyi Liu and Yawei Zhao under the correspondence of Siyu Chen at the Department of Medical Oncology, Xinhua Hospital, Shanghai Jiao Tong University School of Medicine, was built on a strikingly broad computational foundation. The team integrated seven single-cell RNA-sequencing cohorts, two spatial transcriptomic cohorts, bulk RNA sequencing linked to patient survival data, CRISPR-based gene essentiality screens, and three independent immunotherapy cohorts. Rather than relying on any single analytical method, the investigators assembled a machine learning framework in which different algorithms handled different parts of the problem, each feeding its output into the next stage of analysis.

The first step involved unsupervised consensus non-negative matrix factorization, a mathematical technique that decomposes gene expression data into coherent transcriptional programs without any prior labeling. Applied to malignant cells across the cohorts, this approach resolved four distinct malignant meta-programs. One of these, designated MP3, carried a mixed mesenchymal and interferon signaling signature, suggesting tumor cells that had partially undergone epithelial–mesenchymal transition while simultaneously mounting inflammatory responses. Crucially, the researchers found that MP3 carried prognostic information only in the context of the other three programs. When considered in isolation, its signal was misleading; when modeled jointly, patients with higher MP3 activity relative to the other programs fared significantly worse, with a hazard ratio of 1.43 per standard deviation and a p-value of 0.0045.

That context-dependence led the team to construct a composite score, essentially the difference between standardized MP3 activity and the activity of a fourth program, MP4, which resembled healthy alveolar cells and was independently protective. This single composite metric, z(MP3) minus z(MP4), proved prognostic on its own, with a hazard ratio of 1.40 and a p-value of 3.4 multiplied by ten to the negative sixth power. The reciprocal balance between an aggressive, plastic mesenchymal state and a differentiated, alveolar-like state thus encoded clinically meaningful information that neither program revealed alone. It is a technical but important lesson: in tumors defined by cellular plasticity, the relative composition of transcriptional states, not the abundance of any one state, appears to determine patient outcomes.

Having established the malignant programs, the researchers turned to the immune microenvironment and deployed deep generative models to interrogate neutrophils. Using scVI and its supervised extension scANVI, probabilistic models designed to denoise single-cell data and transfer cell type labels across datasets, the team resolved distinct neutrophil states within the tumor microenvironment. What emerged was an OSM-primed neutrophil axis. OSM, or oncostatin M, is an inflammatory cytokine, and neutrophils primed with it appeared to participate in a signaling circuit alongside SPP1-positive macrophages, the very cell population that had dominated prior studies of mesenchymal transition in this cancer. The two cell types formed what the authors describe as a partitioned dual circuit, with neutrophils and macrophages occupying complementary roles in sustaining the mesenchymal niche.

What made this finding particularly compelling was its spatial validation. Single-cell data strips away geography, telling researchers which cells exist but not where they sit within the tumor. To recover that geometry, the team applied cell2location, a probabilistic deep learning method for spatial deconvolution that estimates which cell types and states occupy each spot in a spatial transcriptomics slide. They coupled this with random forest multi-view modeling to map communication fluxes between cell populations. The results showed that OSM signaling flux was directed primarily toward macrophages and fibroblasts rather than toward the malignant cells themselves, which means that any influence of the neutrophil circuit on tumor cells is likely indirect, mediated through the stromal and macrophage compartments. The two-compartment niche, pairing OSM-primed neutrophils with their macrophage and fibroblast partners, was reproducible across both spatial cohorts, strengthening confidence that it reflects genuine tumor architecture rather than computational artifact.

The final and perhaps most ambitious stage of the pipeline used Geneformer, a transformer-based single-cell foundation model pretrained on large corpora of gene expression data, to perform in silico gene deletion perturbations. In effect, the model simulates what happens to a cell’s transcriptional state when a particular gene is removed, allowing researchers to computationally screen candidate dependencies across the three coupled state transitions identified in the study: from alveolar-like to mesenchymal malignant states, and through the associated neutrophil and macrophage circuits. The screen converged on SEC61G, a gene encoding a component of the SEC61 translocon, the protein channel in the endoplasmic reticulum membrane through which secreted and membrane proteins pass as they are synthesized. Because mesenchymal tumor cells and inflammatory immune cells both rely heavily on protein secretion, a translocon dependency is biologically plausible.

The authors are notably careful about how they frame this nomination. SEC61G already had independent published support in lung adenocarcinoma, so the team treats it as a positive control re-derived de novo rather than a wholly new drug target. What they report as genuinely new is that SEC61G’s prognostic signal is independent of 7p11.2 copy number, the chromosomal region in which the gene resides and a region frequently amplified in this cancer. In other words, the poor outcomes associated with high SEC61G expression are not simply a reflection of having more copies of the gene, hinting at regulatory or functional dependencies that copy number analysis alone would miss. The overall survival hazard ratio for SEC61G was 1.64 with a p-value of 1.3 multiplied by ten to the negative sixth, and the gene showed higher expression in immunotherapy non-responders in two of the three checkpoint inhibitor cohorts examined, reaching statistical significance in one.

What elevates this study above a typical bioinformatics exercise is its insistence on orthogonal validation. The four-endpoint validation cascade required convergent evidence from survival analysis, immunotherapy response data, CRISPR essentiality screens, and the perturbation modeling itself before any candidate dependency was accepted. This design directly addresses one of the most persistent criticisms of single-cell oncology research: that computational nominations of drug targets often evaporate under experimental scrutiny. By demanding agreement across data types that share no common analytical machinery, the framework filters out candidates whose signals are artifacts of any single method or dataset. The result is a shorter but far more defensible list of candidate targets than a typical computational screen would produce.

The broader significance lies in the generalizability of the approach. The authors explicitly position their strategy, foundation model perturbation followed by multi-endpoint orthogonal validation, as a reusable template for nominating therapeutic targets in plasticity-driven solid tumors, a category that includes many of the hardest cancers to treat. As foundation models like Geneformer mature and spatial transcriptomic datasets accumulate, pipelines of this kind could compress the path from observational single-cell atlases to testable therapeutic hypotheses. For lung adenocarcinoma patients, the immediate deliverables are a newly mapped neutrophil-associated mesenchymal niche that reframes how the tumor microenvironment sustains aggressive cell states, and a candidate translocon dependency whose vulnerability can now be pursued with experimental tools. The work was funded by the National Natural Science Foundation of China and the Shanghai Committee of Science and Technology, and relied exclusively on publicly available, de-identified human data, meaning that its findings can be independently reanalyzed and challenged by any laboratory with computational resources and internet access.

Subject of Research: Deep learning integration of single-cell and spatial transcriptomics to map a neutrophil-associated mesenchymal niche and identify candidate tumor dependencies in lung adenocarcinoma.

Article Title: Deep learning integration of single-cell and spatial transcriptomics reveals a neutrophil-associated mesenchymal niche and a candidate translocon dependency in lung adenocarcinoma

Article References: Deep learning integration of single-cell and spatial transcriptomics reveals a neutrophil-associated mesenchymal niche and a candidate translocon dependency in lung adenocarcinoma. (n.d.). https://doi.org/10.1007/s00262-026-04560-3

Image Credits: AI Generated

DOI: 10.1007/s00262-026-04560-3

Keywords: lung adenocarcinoma, single-cell RNA sequencing, spatial transcriptomics, tumor-associated neutrophils, epithelial–mesenchymal transition, SEC61G, deep learning, tumor microenvironment, Geneformer, immunotherapy response, OSM signaling, cancer prognosis