The study, led by Giwon Kim and Xu Jiang of Kyung Hee University under the supervision of Ki-Hong Jung, appears in the journal Plant Molecular Biology. The researchers profiled the transcriptomes, the complete set of expressed genes, of rice flag leaves at five distinct reproductive stages: one week before heading, at heading itself, and one, three and five weeks after heading. Rather than simply cataloging genes that changed over time, they applied a stringent intersection-based criterion to identify stage-preferential genes, meaning genes whose expression peaked at one particular stage relative to all the others. This approach allowed them to isolate the specific transcriptional signatures of each moment in the grain-filling period rather than blurring them into a single generic aging profile.

The results revealed two dramatic peaks of transcriptional reprogramming. At heading, the moment when the panicle emerges and flowering begins, 419 genes were preferentially expressed. Gene Ontology analysis showed that these genes were strongly enriched in jasmonate-related processes and defense-associated functions. Jasmonic acid is a plant hormone best known for its roles in wound responses and pest resistance, but previous work has also implicated it in spikelet development and reproductive timing. The new data suggest that as the rice plant transitions from vegetative growth to reproduction, its flag leaf mounts a coordinated hormone- and defense-linked transcriptional program, perhaps protecting the reproductive structures at a moment of exceptional vulnerability.

The second, and far larger, wave of reprogramming came five weeks after heading, when 1,317 genes showed preferential expression. This late-stage gene set told a very different story. Enrichment analysis linked these genes to senescence, nutrient transport, alternative respiration and plastid regulation. In other words, five weeks after heading is when the flag leaf appears to commit fully to its dismantling program: chloroplast components are broken down, transporters are mobilized to move nitrogen, phosphorus and mineral ions out of the leaf, and metabolic pathways shift toward the catabolic reactions that convert cellular infrastructure into exportable nutrients. Alternative respiration, a mitochondrial pathway that can help manage the reactive oxygen species generated during cellular breakdown, also featured prominently, hinting at how the senescing leaf keeps its energy metabolism functional even as its photosynthetic machinery is dismantled.

To move from a list of stage-preferential genes to an understanding of functional relationships, the team turned to network biology. They used the STRING database, which compiles known and predicted protein-protein associations, to construct an interaction network combining the genes preferentially expressed at five weeks after heading with reference senescence genes drawn from the Leaf Senescence Database. Within the largest connected component of this network, they applied a maximal clique centrality algorithm to identify hub candidates, the most highly connected and presumably most influential proteins in the system. The hubs that emerged clustered around two major biological themes: chloroplast and chlorophyll turnover, and nitrogen remobilization. This convergence is biologically telling, because the chloroplast holds the majority of the leaf’s nitrogen in the form of photosynthetic proteins, so breaking down chlorophyll-protein complexes is simultaneously the visible hallmark of senescence and the engine of nitrogen export.

Among the nitrogen-related hub candidates were components long associated with glutamine synthetase activity, the enzymatic gateway through which organic nitrogen is prepared for transport out of the leaf. Decades of research, from early work on glutamine synthetase in naturally senescing rice leaves to recent studies of cytosolic glutamine synthetase isoforms in grain ripening, have established this pathway as central to yield formation. By anchoring it within a stage-resolved interaction network, the new study provides a prioritized shortlist of genes that breeders and molecular biologists can now interrogate as potential levers for improving nitrogen use efficiency, a trait of enormous agronomic and environmental importance given the costs and consequences of nitrogen fertilizer.

The researchers also validated their findings experimentally using quantitative RT-PCR in the pale-green leaf mutant, or pgl, a rice line carrying a defect in a gene encoding chlorophyllide a oxygenase 1, an enzyme involved in chlorophyll metabolism. The pgl mutant is known to senesce differently from wild-type plants, indirectly affecting grain yield and quality. When the team measured the expression of representative hub genes in the mutant, they found broadly reduced expression compared with the wild type. This result suggests that the late-stage senescence network identified at five weeks after heading is attenuated in pgl, providing an independent line of evidence that the hub genes are genuine functional components of the senescence program rather than statistical artifacts of the profiling pipeline.

Technically, the study exemplifies the modern toolkit of plant genomics. RNA sequencing data were processed with standard trimming and alignment pipelines, expression quantified as transcripts per million, differential expression assessed with rigorous false discovery rate control, and time-course patterns examined with dedicated clustering methods. Data visualization relied on heatmap frameworks, and functional interpretation drew on enrichment tools and metabolic mapping platforms. The complete RNA-seq dataset has been deposited in ArrayExpress at EMBL-EBI under accession number E-MTAB-16817, making the resource freely available to the research community. For a crop that feeds more than half the world’s population, an open, stage-resolved reference of flag leaf biology is a contribution that extends well beyond the individual laboratory that generated it.

What makes the study particularly valuable is its insistence on stringency. Many transcriptomic surveys of leaf senescence collapse time points into broad categories and report long lists of differentially expressed genes, leaving researchers to guess which candidates matter most. By requiring that a gene be upregulated at one stage relative to every other stage, the Korean team produced a far more restrictive gene set, one in which each entry carries a clear temporal identity. The two major hubs of activity, heading and five weeks after heading, now stand as well-defined windows in which breeders might look for natural variation or in which genome editors might intervene. Delaying the late senescence program slightly, for example, could extend the photosynthetic duration of the flag leaf, while enhancing the remobilization program could improve the efficiency with which nutrients reach the grain.

The work also reframes an old question in crop physiology. Scientists have long known that a rice plant’s last leaf is both its factory and its warehouse, and that the timing of the warehouse’s liquidation is a matter of delicate balance. Senesce too early, and the grain is starved of photosynthate; senesce too reluctantly, and nutrients remain locked in the leaf. By showing that the transition is governed by discrete, stage-specific transcriptional programs, one defensive and hormonal at heading, one catabolic and export-oriented five weeks later, the study offers a molecular vocabulary for describing that balance. It links visible yellowing to specific network modules, and those modules to testable candidate genes. As global rice production faces mounting pressure from climate variability and the need to reduce fertilizer inputs, understanding the genetic timetable of the flag leaf may prove to be one of the more consequential stories in modern plant science, told this time gene by gene, stage by stage, in the fading green of a single leaf.

Subject of Research: Stage-preferential transcriptome profiling of rice flag leaf senescence and nutrient remobilization during grain filling.

Article Title: Stage-preferential transcriptome profiling reveals senescence associated transcriptional programs linked to nutrient remobilization in rice flag leaves during grain filling

Article References: Kim, G., Jiang, X., Yoo, Y.-H., Hong, W.-J., & Jung, K.-H. (2026). Stage-preferential transcriptome profiling reveals senescence associated transcriptional programs linked to nutrient remobilization in rice flag leaves during grain filling. Plant Molecular Biology, 116(5), Article 98. https://doi.org/10.1007/s11103-026-01762-2

Image Credits: AI Generated

DOI: 10.1007/s11103-026-01762-2

Keywords: rice flag leaf, grain filling, leaf senescence, transcriptome, nutrient remobilization, pale-green leaf mutant, protein-protein interaction network, chlorophyll degradation, nitrogen remobilization, jasmonic acid, RNA-seq, Plant Molecular Biology

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