The research, published in Frontiers in Cardiovascular Medicine on 20 October 2026, included 107 adults treated at Shanghai Changzheng Hospital between January 2023 and January 2026. All participants had migraine diagnosed under International Classification of Headache Disorders criteria and a patent foramen ovale with right-to-left shunting confirmed by cardiac imaging. The closure group contained 58 patients who received a transcatheter occluder, while the control group comprised 49 patients who received standard medical therapy without closure. Follow-up assessments covered headache episodes, pain intensity, migraine impact, disability, quality of life and depressive symptoms at one week and at 1, 3, 6, 9 and 12 months. Overall follow-up was complete through six months, with 105 of 107 patients, or 98.13%, assessed at 12 months.

The groups had similar distributions of sex, age and several recorded clinical characteristics, but important differences remained. Every patient in the closure group had a moderate or large right-to-left shunt: 38 were classified as large and 20 as moderate. The control group had 37 small and 12 moderate shunts and no large shunts. The closure group also began with greater headache severity, averaging 6.20 on the Visual Analogue Scale compared with 4.59 in controls. Because these baseline differences could affect raw follow-up scores, the investigators compared pre-to-post changes and performed multivariable regression adjusting for age, sex, baseline shunt grade and migraine subtype. Migraine with aura was present in 38 of 58 closure patients and 32 of 49 controls.

Headache severity improved more in the closure group at every follow-up assessment. The mean reduction in Visual Analogue Scale score was 3.28 points at one week in the closure group, compared with 0.92 points in controls. At 12 months, the corresponding reductions were 3.48 and 0.94 points, with P values below 0.001 for the intergroup comparisons at all measured time points. The closure group also had greater changes in Headache Impact Test, Migraine Disability Assessment, Migraine-Specific Quality of Life Questionnaire and Beck Depression Inventory-II scores, with every reported intergroup comparison yielding P values below 0.001. In the adjusted model, closure was associated with 2.86 additional points of 12-month Visual Analogue Scale improvement, with a 95% confidence interval of 1.92 to 3.80, after accounting for the measured covariates.

Patients with large shunts appeared to have greater improvement than those with moderate shunts. At 12 months, Visual Analogue Scale improvement averaged 4.12 points among the 38 closure patients with large right-to-left shunts and 2.86 points among the 20 with moderate shunts, a difference with P = 0.032. This subgroup observation included only 20 participants in the moderate-shunt category and had no corresponding control-group comparison, so it does not establish that shunt size determines who benefits. The study’s efficacy categories were based primarily on Visual Analogue Scale change: a reduction of at least five points or a final score of zero was classified as markedly effective, any reduction without reaching those thresholds as effective, and no reduction or an increase as ineffective. At 12 months, 37 closure patients were markedly effective, 18 were effective and three were ineffective, compared with 10, 26 and 13, respectively, in controls. The markedly effective and ineffective distributions differed significantly, while the difference in the effective category did not.

Blood-gas measurements in the closure group improved after the procedure. One month after closure, mean arterial partial pressure of oxygen, PaO₂, had risen from 82.86 to 85.84 mmHg, while mean arterial oxygen saturation, SaO₂, increased from 95.28% to 96.69%. Both changes were statistically significant, with P values below 0.01. Contrast-enhanced echocardiography performed one month after all 58 procedures showed complete closure of the foramen ovale without residual right-to-left shunting. The investigators considered the oxygenation findings consistent with elimination of intracardiac shunting, but the study did not provide an equivalent pretreatment blood-gas comparison for the medical-therapy group. The measurements therefore establish a temporal association within the closure group rather than demonstrating a causal pathway from shunt correction to headache improvement.

For the molecular analysis, the researchers selected five closure-group patients whose migraine had completely resolved after the procedure and compared their preoperative and one-month postoperative arterial plasma with samples from five age- and sex-matched healthy volunteers. Tandem mass-tag labeling and liquid chromatography-tandem mass spectrometry identified approximately 790 proteins. Before closure, the patient samples separated from healthy controls in principal-component analysis, and the preoperative and postoperative patient samples also showed some separation. Eighteen proteins met the study’s differential-expression threshold when preoperative and postoperative patient samples were compared. Myeloperoxidase, lysozyme C and heat shock protein 60 were described as inflammation- and oxidative-stress-related proteins whose expression was lower after closure; the findings do not demonstrate a direct reduction in oxidative stress itself.

Gene Ontology analysis of proteins that differed before and after closure identified enrichment in inflammatory-response regulation, response to oxidative stress, bacterial defense, gluconeogenesis and cellular responses to amino acids. Cellular annotations included extracellular exosomes, extracellular space, azurophil granule lumens and secretory granules, which the authors associated with immune-cell activation. Postoperative samples also showed relative increases in proteins involved in vascular or extracellular-matrix functions, including angiotensin-converting enzyme, collagen alpha-1(VI) chain and integrin alpha-2. Untargeted metabolomics identified differences in dopamine, L-proline, L-tyrosine, D-proline, acetylcarnitine and galactitol between post-closure patients and healthy controls, with pathway signals involving protein digestion and absorption, metabolism, transporter activity, and arginine and proline metabolism. These exploratory signals were not independently tested as predictive biomarkers, and the study did not show that they mediated the observed headache improvement.

Procedure-related outcomes were reassuring but remain limited by the small sample. Eight patients reported palpitations after activity and two reported postoperative chest tightness. Two patients developed new-onset atrial fibrillation, an incidence of 3.4%; both spontaneously returned to sinus rhythm and had no recurrence during the 12-month study period. No cardiac tamponade, device migration or embolization, thrombosis, infection or arteriovenous fistula occurred. The study’s principal limitations were its retrospective, single-center design, treatment-related differences in shunt severity and baseline symptoms, differing age eligibility ranges, reliance on self-reported headache outcomes and incomplete long-term imaging follow-up. The proteomic analysis included only five patients and five controls, healthy controls were not all screened by transesophageal echocardiography, and the direction of causality between the molecular changes, oxygenation and symptoms remains unresolved. The authors therefore describe the results as preliminary and say that larger prospective cohorts and randomized trials are needed before PFO closure can be considered a standard treatment for migraine or the candidate proteins can guide patient selection.

Subject of Research: Clinical outcomes and exploratory blood-protein changes associated with patent foramen ovale closure in patients with migraine

Article Title: Preliminary exploration of the pathogenesis of patent foramen ovale-related migraine and efficacy analysis of PFO closure

Article References: Ji, M., Xu, J., Luo, C., Gao, M., Pu, W., Peng, H., & Xiao, J. (2026). Preliminary exploration of the pathogenesis of patent foramen ovale-related migraine and efficacy analysis of PFO closure. Frontiers in Cardiovascular Medicine, 13, Article 1928840. https://doi.org/10.3389/fcvm.2026.1928840

Image Credits: AI Generated

DOI: 10.3389/fcvm.2026.1928840

Keywords: patent foramen ovale, migraine, transcatheter closure, right-to-left shunt, retrospective cohort, plasma proteomics, arterial oxygenation, migraine with aura, inflammation, oxidative stress, Preliminary, exploration

The trial in question, reported by Alex Zhavoronkov and colleagues in Nature Biotechnology, marks the first time that six distinct proteomic aging clocks have been deployed together as readouts in a clinical study of an anti-aging intervention. That detail matters because clinical trials are unforgiving environments for biomarkers. A clock that looks impressive when trained on cross-sectional blood samples from tens of thousands of people may behave very differently when asked to detect a modest, treatment-induced change within individuals over months. Robertson and Chandra’s verdict on the exercise is pointed: the deployment revealed more about the clocks themselves than about the drug being tested. In other words, the trial functioned less as a decisive test of a therapy and more as a stress test of the measurement instruments.

To understand why this matters, it helps to consider how proteomic clocks are built. Researchers start with large cohorts in which plasma protein levels have been measured, typically using affinity-based platforms that quantify thousands of proteins simultaneously. Statistical models, often penalized regressions or machine-learning approaches, are then trained to predict chronological age from those protein patterns. The difference between the predicted age and a person’s true age, the so-called age gap, is interpreted as a measure of biological aging. Large-scale efforts published in recent years, including analyses of UK Biobank data by Argentieri and colleagues in Nature Medicine in 2024 and independent work by Xu and colleagues in 2025, showed that such proteomic clocks predict mortality and incident disease with genuine skill, in some cases capturing organ-specific deterioration that epigenetic measures miss.

But prediction in a population is not the same as sensitivity to change in an individual. A clock can be highly accurate at ranking people by age and still be nearly useless for detecting whether a six-month intervention slowed anyone’s aging. Robertson and Chandra frame the trial through this distinction, which echoes a long-standing concern in clinical medicine formalized by Fleming and DeMets in a 1996 paper in Annals of Internal Medicine: a surrogate endpoint must capture the effect of treatment on the outcome of interest, not merely correlate with the outcome in observational data. Correlation across individuals does not guarantee that a biomarker will respond, or respond in the right direction, when a drug perturbs the underlying biology. This is the central technical challenge now facing proteomic geroscience.

The trial also exposed subtler problems that arise when multiple clocks are run on the same samples. Different proteomic clocks are trained on different platforms, different cohorts and different sets of proteins, and they can disagree. Two clocks might both predict chronological age well, yet one might show a treatment effect while another shows none, or even a shift in the opposite direction. Without a validated gold standard, it is difficult to know which clock, if any, is tracking true biological aging rather than platform noise, inflammation from an intercurrent illness, or compositional changes in blood cell populations. Robertson and Chandra emphasize that the six-clock comparison turned this ambiguity into data, allowing the field to see where the clocks converged and where they diverged when confronted with real longitudinal samples from a controlled study.

There is also a deeper statistical issue lurking beneath the age-gap framework. Proteomic clocks are trained to predict chronological age, so by construction the model tends to pull predictions toward the population mean. If a treatment genuinely shifts a person’s protein profile toward a younger state, the clock should register a negative age gap. But the same shift can be produced by regression to the mean, by changes in the overall variance of the proteome, or by the treatment altering the levels of a handful of highly weighted proteins without broadly rejuvenating the system. Dabrowski and colleagues explored related questions in Nature Aging in 2024, examining how clock construction choices shape what the resulting scores actually measure. The trial setting, with its randomized structure and repeated sampling, offers a way to interrogate these artifacts directly rather than speculate about them.

The stakes are high because the geroscience field has been waiting for credible endpoints. Interventions such as rapamycin, senolytic drugs and metformin are being tested in humans with the aim of extending healthspan, yet aging itself is not an approved trial endpoint. The TARGET trial of dasatinib and quercetin, reported by Justice and colleagues in Geroscience in 2018, was an early attempt to pair a senolytic intervention with biomarker readouts, and Moqri and colleagues laid out a systematic framework for biomarker validation in a 2023 review in Cell. What has been missing is the bridge between those frameworks and actual regulatory-grade evidence. A clock that reliably detects decelerated aging in a trial would transform the economics of the field, because trials measuring aging directly could be far shorter and smaller than trials waiting for fractures, infections or deaths to accumulate.

The commentary’s sober conclusion, that the trial taught the field more about the clocks than about the drug, should be read as progress rather than disappointment. First-in-class instruments are always refined through use. The trial provided longitudinal, randomized data against which clock behavior could be benchmarked, something no retrospective cohort can offer. It likely revealed issues of test-retest reliability, sensitivity floors and the influence of acute physiological fluctuations on plasma proteins, all of which are invisible in cross-sectional training data. Robertson and Chandra suggest that this kind of head-to-head deployment, with multiple clocks applied to the same trial samples, is exactly what the field needs to converge on which proteomic signatures are robust enough to carry forward toward validation as surrogate endpoints.

The path forward, as the commentary implies, will require standardization and skepticism in equal measure. Proteomic platforms need consistent calibration across laboratories; clocks need pre-registration and locked algorithms before trials begin, so that researchers cannot select the readout that happens to favor their result; and the field needs studies that test clocks against hard clinical outcomes prospectively, not just against chronological age. The 2026 trial of Zhavoronkov and colleagues will likely be remembered less for any verdict on its experimental therapy and more as the moment proteomic aging clocks left the training ground. If the lessons learned here are absorbed, the next generation of clocks may enter trials not as optimistic proxies but as instruments with documented performance characteristics, bringing the dream of measuring, and eventually modulating, human aging a measurable step closer.

Subject of Research: Validation of proteomic aging clocks as biomarkers in a clinical trial

Article Title: Proteomic aging clocks go to trial

Article References: Robertson, N. A., & Chandra, T. (2026). Proteomic aging clocks go to trial. Nature Biotechnology. https://doi.org/10.1038/s41587-026-03361-4

Image Credits: AI Generated

DOI: 10.1038/s41587-026-03361-4

Keywords: proteomic aging clocks, biological age, biomarkers, clinical trial, geroscience, plasma proteomics, surrogate endpoints, epigenetic clocks, longevity research, Nature Biotechnology, age prediction, biomarker validation

The scale and clinical breadth of the underlying cohort is what sets the resource apart. Most transcriptomic studies focus on one disease at a time, comparing patients against a small control group, which makes it difficult to know whether a gene signature is truly specific to a condition or merely reflects generic illness. By collecting whole blood RNA profiles across 98 different health states within a coordinated framework, the researchers could ask, for the first time at this scale, which molecular changes are shared across many diseases and which are genuinely distinctive. The atlas characterizes each condition at multiple levels of biological organization: individual genes, co-expressed gene modules, and entire biological pathways, with particular attention to immune cell composition, metabolic activity, and cell-type-specific signals that can be inferred from bulk blood measurements.

Technically, whole blood transcriptomics presents a deconvolution problem. A blood sample is a mixture of red cells, platelets, and a diverse population of leukocytes, including neutrophils, monocytes, lymphocytes, and dendritic cells, each with its own expression program. When a disease shifts the proportions of these cell types, the bulk RNA profile changes even if no individual cell alters its internal state. The atlas addresses this by explicitly modeling immune and cell-type-specific features alongside gene and pathway-level signals, allowing the authors to separate changes driven by cellular composition from changes driven by transcriptional regulation within cells. This layered characterization means that a researcher probing, say, a suspected autoimmune condition can distinguish whether the dominant signal is an expansion of neutrophil-associated transcripts or a genuine upregulation of interferon-response genes inside lymphocytes.

A second major component of the study extends the atlas beyond RNA. For 1,280 participants representing 28 of the health conditions, the team generated matched plasma proteomics data, measuring the circulating protein landscape in the same individuals whose blood RNA had been profiled. This pairing allowed a systematic assessment of the relationship between messenger RNA abundance in blood cells and protein concentration in plasma, a relationship that is notoriously imperfect. Many transcripts are translated at low efficiency, many proteins are secreted by tissues far from the bloodstream, and many blood proteins have half-lives that decouple them from the transient RNA signals that produced them. By quantifying these RNA-protein relationships across dozens of conditions, the researchers demonstrated that the two molecular layers carry complementary rather than redundant information, each capturing aspects of disease biology the other misses.

That complementarity became the foundation for the study’s most translational result: multi-class classification models that integrate RNA and protein signatures to distinguish among a wide range of health conditions. Where most diagnostic machine learning studies tackle a binary question, diseased versus healthy, the atlas models were trained to assign samples to the correct condition among many alternatives, a substantially harder task that mirrors real clinical differential diagnosis. The integrated multi-omics models consistently outperformed classifiers built on RNA data alone or protein data alone, confirming quantitatively that the information content of the combined signatures exceeds the sum achievable with either layer by itself. The authors validated the identified RNA and protein signatures across ten independent whole blood transcriptomics cohorts and in the UK Biobank, providing external replication that is rare in single-center omics studies.

The validation strategy deserves emphasis because it speaks to a chronic weakness of the biomarker literature. Candidate signatures discovered in one cohort frequently fail when tested elsewhere, a phenomenon driven by differences in sequencing platforms, sample handling, population demographics, and disease prevalence. By requiring that the atlas-derived signatures hold up in ten external transcriptomics datasets and in the population-scale UK Biobank resource, used under Application Number 150710, the team provided evidence that the disease-associated patterns are robust biological signals rather than artifacts of a particular laboratory or cohort. For clinicians and translational researchers, this kind of cross-cohort reproducibility is the difference between an interesting observation and a candidate diagnostic tool.

The consortium behind the work is itself notable. Alongside the Stockholm computational team, which includes Mathias Uhlen, Saeed Shoaie, Cheng Zhang, and corresponding author Adil Mardinoglu, the author list spans dozens of clinician-researchers at Atatürk University’s Faculty of Medicine, Erzurum Regional Training and Research Hospital, and partner institutions, reflecting the enormous clinical recruitment effort required to assemble thousands of well-phenotyped patient samples across such a wide diagnostic spectrum. The work received support from the Knut and Alice Wallenberg Foundation, and the resulting atlas is being positioned as an open resource for the precision medicine community, enabling other groups to query disease-associated transcriptomic signatures without repeating the costly sample collection.

What might the atlas enable in practice? For diagnostic developers, the multi-class models suggest a route toward blood tests that do not merely flag that something is wrong but help triangulate what is wrong, potentially accelerating workups in settings where patients present with nonspecific symptoms. For drug developers, the pathway and module-level maps offer a way to see which biological processes are perturbed across disease families, highlighting shared inflammatory or metabolic axes that might be targeted therapeutically or that explain why a single drug produces effects across seemingly unrelated conditions. For basic researchers, the resource provides a reference against which new patient data can be projected, instantly contextualizing an individual expression profile within a landscape of 98 known health states.

The study also contributes a methodological lesson about the biology of blood itself. The finding that RNA and protein layers play complementary biological roles underscores that neither measurement alone can serve as a complete proxy for physiological state. Transcripts reflect what blood cells are actively doing at the moment of sampling; plasma proteins reflect secretion, clearance, and contributions from distant organs. Diseases leave different imprints on each layer depending on their mechanism, whether they act primarily through immune activation, tissue damage, or systemic metabolic change. An atlas that records both layers, in matched samples, across a broad disease spectrum, therefore functions less like a single dataset and more like a bilingual dictionary for translating between cellular activity and circulating physiology.

Published as an open-access article with extensive supplementary datasets, the WBT Atlas arrives at a moment when large-scale, multi-omic reference resources are reshaping how biomedical research is organized, echoing the impact of earlier protein and tissue atlases from the same Stockholm research environment. Its authors describe it as a genome-wide resource enabling translational discovery in precision medicine, and the external validations included in the paper suggest the description is more than aspiration. As more cohorts and conditions are added and as the community begins projecting new patient data onto the map, the atlas may well become a standard reference point, the molecular backdrop against which the next generation of blood-based diagnostics is drawn.

Subject of Research: A whole blood transcriptomics atlas characterizing disease-associated gene, module, and pathway signatures across 98 human health conditions with integrated plasma proteomics

Article Title: Whole blood transcriptomics atlas across human diseases

Article References: Li, M., Tatar, A., Green, E., Simsek, E., Kahraman, C. Y., Ceylan, M., Laloglu, F., Ceviz, N., Tatar, A., Laloglu, E., Akcay, G., Cadirci, K., Yapca, O. E., Ozmen, H. K., Yapanoglu, T., Kudas, O., Simsek, F., Yılmaz, E. P. T., Senocak, G. N. C., … Mardinoglu, A. (2026). Whole blood transcriptomics atlas across human diseases. Nature Communications. https://doi.org/10.1038/s41467-026-78440-w

Image Credits: AI Generated

DOI: 10.1038/s41467-026-78440-w

Keywords: transcriptomics, whole blood, biomarkers, multi-omics, plasma proteomics, precision medicine, machine learning, gene expression, disease classification, UK Biobank, Nature Communications, diagnostics

The study, published in Aging Cell, is among the first to benchmark eleven proteomic organ clocks against molecular, imaging, cognitive and physical biomarkers within a single population. The organ clocks were originally developed by mapping thousands of plasma proteins to their tissues of origin and training mathematical models to predict chronological age for organs including the brain, heart, liver, kidney, immune system and arteries. When the researchers applied these clocks to the Scottish cohort, they found substantial heterogeneity: organs within the same person appeared to age at strikingly different rates, and the organ age gaps were only mildly correlated with one another, with a mean pairwise correlation of 0.32. Extreme ageing, defined as an organ age gap exceeding two standard deviations from the population mean, most commonly affected just a single organ rather than several at once.

Despite this biological richness, the organ clocks delivered only modest prognostic power. Each standard deviation of accelerated organ ageing corresponded to a 16 to 43 percent higher mortality hazard over the follow-up period. The strongest effects emerged for the liver, with a hazard ratio of 1.43, followed by the immune system at 1.42 and the heart at 1.38. The proteomic brain age gap, at a hazard ratio of 1.31, performed almost identically to a brain age gap derived from magnetic resonance imaging, which scored 1.30. These findings replicate the original organ clock study, which reported a 17 to 53 percent mortality increase per standard deviation in a separate American cohort, confirming that the approach generalises across populations.

Yet several other biomarkers left the organ clocks behind. The clear champion was GrimAge2, a second-generation epigenetic clock that estimates biological age from DNA methylation patterns in blood. A single standard deviation of GrimAge2 acceleration conferred a 62 percent higher mortality hazard, the largest effect of any biomarker tested. Close behind came structural brain measures: smaller total brain volume raised mortality risk by 52 percent per standard deviation, and smaller grey matter volume by 44 percent. Reduced lung function also proved formidable, with a one-second forced expiratory volume deficit carrying a 51 percent higher hazard, and poorer general cognitive function adding 46 percent. All of these surpassed every proteomic organ clock in effect size.

The researchers then asked which biomarkers carried independent information when all were considered simultaneously. In a multivariable Cox model including 21 ageing measures among the 460 participants with complete data, only four survived: total brain volume, white matter hyperintensity volume, general cognitive function and walking time. GrimAge2, despite its stellar univariable performance, lost significance, not because it lacks biological relevance but because much of its prognostic signal overlaps with correlated measures. Together, the four independent biomarkers explained roughly 19 percent of the variance in mortality risk, while all 21 biomarkers combined raised that figure only to 23 percent, suggesting diminishing returns from adding further measures.

The variance decomposition was equally revealing. Chronological age and sex accounted for just 2 percent of mortality risk, with lifestyle factors adding 4 percent. GrimAge2 contributed a further 6 percent, proteomic organ clocks 8 percent and physical function measures 9 percent. Brain-related measures, combining neuroimaging and cognition, accounted for around 16 percent, roughly double any other single domain. For the authors, this underscores that brain health, encompassing atrophy, cerebrovascular damage and cognitive capacity, is a particularly powerful window on survival in old age, likely because the brain regulates essential physiological functions and because atrophy captures neurodegenerative processes long before clinical diagnosis.

Beyond benchmarking clocks, the team mined the full plasma proteome for mortality signals. Using the SomaScan 11K platform, which quantifies nearly 10,000 protein targets from a small plasma sample, they identified 368 proteins significantly associated with all-cause mortality after adjustment for age and sex. The strongest positive associations belonged to GDF15, a stress- and inflammation-induced cytokine, with a hazard ratio of 1.56, followed by WFDC2 at 1.47 and TIMP1 at 1.45. Notably, GDF15 ranked as the second strongest mortality predictor overall, behind only GrimAge2, and three of the nine proteins whose DNA methylation surrogates are built into GrimAge2 appeared among the top twenty. On the protective side, higher circulating levels of neuropeptide S, BAGE3 and ABCC6 each corresponded to roughly 41 to 42 percent lower mortality risk per standard deviation.

Gene set enrichment analysis revealed a striking functional dichotomy among these proteins. Those whose higher levels predicted death were overwhelmingly enriched for immune processes, including antimicrobial humoral immune responses, chemokine and cytokine activity, and interleukin-4 and interleukin-13 signalling, painting a picture of inflammaging and immunosenescence, the chronic low-grade inflammation and immune decline that characterise ageing. Proteins associated with survival, by contrast, were enriched for genomic integrity: nucleosome assembly, regulation of the G2/M cell cycle transition, damaged DNA binding and ribosome biogenesis. Among the protective candidates were Sirtuin-2 and Caspase-2, both involved in DNA damage response and tumour suppression, with higher levels linked to 18 and 32 percent lower mortality risk respectively. An elastic net penalised regression across all 9,703 protein targets distilled the signal to just 37 proteins, with GDF15 carrying the largest positive coefficient and strong concordance with the univariate results.

The study has limitations the authors acknowledge. The cohort consists of relatively healthy, ethnically homogeneous Scots born in 1936, so validation in more diverse populations is needed. The analyses relied on single baseline measurements rather than longitudinal trajectories, examined all-cause rather than cause-specific mortality, and depended on one proteomics platform. The findings are observational and cannot establish causality. Still, the practical implications are considerable: simple, inexpensive measures such as walking speed, lung function and cognition rival or exceed costly molecular assays, while integrative biomarkers trained directly on mortality outcomes, like GrimAge2, currently offer the strongest single-molecule performance. Whether organ clocks add value for cause-specific mortality, such as distinguishing cardiac from neurological deaths, remains an open question.

What emerges is a clear message for the booming biological-ageing industry: no single number captures how fast someone is ageing. Organs age at their own pace, blood proteins tell one story, brain scans another, and the most informative forecast comes from combining complementary measures across modalities. As the world’s population over 60 doubles to a projected 2.1 billion by 2050, choosing the right combination of biomarkers could determine who benefits from early intervention, and the new results suggest that protecting the brain, preserving lung capacity and taming chronic inflammation may matter more than any single molecular clock.

Subject of Research: Multimodal ageing biomarkers and plasma proteomic signatures associated with all-cause mortality in older adults

Article Title: Multimodal Ageing Biomarkers and Plasma Proteomic Signatures Associated With All‐Cause Mortality

Article References: Pyrgioti, M., Eguiagaray, I. M., Redmond, P., Corley, J., Bastin, M. E., Hernández, M. V., Russ, T. C., Wardlaw, J. M., Hannon, E., Deary, I. J., Walker, K. A., Tucker‐Drob, E. M., Cox, S. R., Marioni, R. E., & Harris, S. E. (2026). Multimodal Ageing Biomarkers and Plasma Proteomic Signatures Associated With All‐Cause Mortality. Aging Cell, 25(10), Article e70747. https://doi.org/10.1111/acel.70747

Image Credits: AI Generated

DOI: 10.1111/acel.70747

Keywords: biological age, proteomic organ clocks, epigenetic clock, GrimAge2, all-cause mortality, plasma proteomics, GDF15, brain atrophy, cognitive function, lung function, immunosenescence, Lothian Birth Cohort 1936

The modification at the heart of the study is deceptively simple. Lysine is one of the twenty amino acids that make up proteins, and when a small acetyl group is attached to its side chain, the physics of the protein changes. Acetylation can neutralize a positive charge, alter how the protein folds, reshape the surfaces it uses to bind partners, change how long it survives in circulation, and modulate its enzymatic activity. Because a single protein can carry acetyl groups at many different lysine positions, each with potentially distinct consequences, the modification multiplies the functional diversity of the proteome many times over. In cells and tissues, lysine acetylation is famous as a regulator of gene expression and metabolism; in blood plasma, the protein-rich, cell-free fraction of blood, it has been far less systematically explored, particularly in the context of neurodegenerative disease.

To capture this hidden layer, the researchers turned to mass spectrometry, the workhorse technology of modern proteomics. In this approach, proteins extracted from plasma are broken into short peptide fragments, which are then separated by liquid chromatography and fired into a mass spectrometer. The instrument fragments each peptide further and records the resulting spectra, from which the identity of the parent protein and the exact position of any chemical modification can be inferred. Crucially, the team enriched specifically for acetylated peptides before analysis, a step that is essential because acetylated peptides are typically rare amid the overwhelming abundance of unmodified material in plasma. Without such enrichment, the signal from these modified peptides would drown in noise.

The scale of the resulting dataset is what sets the study apart. From plasma samples obtained from twenty Alzheimer’s disease patients and twenty age-matched controls, the researchers acquired more than 7,600 independent tandem mass spectrometry spectral measurements corresponding to 350 unique lysine acetylation sites. Each of those sites represents a specific lysine residue on a specific plasma protein where an acetyl group was detected and localized. This is not merely a catalog of proteins present in blood, which earlier plasma proteomics efforts have assembled in depth, but a map of where on each protein the chemical decoration occurs, in a clinical context where such information did not previously exist.

The study is, according to its authors, the first lysine acetylation-directed proteomic landscape of Alzheimer’s-related human plasma samples. That first matters because plasma is where blood-based diagnostics for Alzheimer’s are increasingly being developed. Current leading biomarkers, such as measures of phosphorylated tau in plasma, focus on a different post-translational modification, phosphorylation, and on proteins that originate in the brain. Acetylation offers a complementary view: it reports on the state of the abundant plasma proteome itself, including lipoproteins, complement factors, coagulation proteins and carriers that mediate inflammation, lipid transport and immune signaling, all of which are increasingly implicated in Alzheimer’s pathogenesis through large-scale genetic and epidemiological studies.

Among the proteins carrying acetylation marks, apolipoproteins stand out as a biologically plausible focus. These lipid-binding proteins shuttle cholesterol and other fats through the bloodstream, and the strongest genetic risk factor for late-onset Alzheimer’s disease, the APOE gene, encodes an apolipoprotein. The study’s abbreviation list also flags low-density lipoprotein, the cholesterol-carrying particles whose biology intersects with both cardiovascular and neurodegenerative disease, suggesting that acetylation states on lipid transport machinery were among the features captured in the dataset. How acetylation alters the function of these carriers, and whether specific acetylation patterns track with disease status, are questions the dataset now makes answerable.

Perhaps the most unexpected dimension of the work lies beyond canonical proteins altogether. The mass spectrometry measurements detected acetylated peptide sequences that do not map to the standard catalog of human proteins. Instead, they correspond to short open reading frames, stretches of the genome that ribosome profiling technologies have shown are actively translated into tiny peptides even though they were long dismissed as non-coding. The idea that such non-canonical peptides circulate in plasma, and that they carry lysine acetylation marks detectable by mass spectrometry, expands the searchable space of blood-borne molecules. It hints that the plasma proteome is even less fully inventoried than assumed, and that some disease-relevant signals may reside in proteins that standard annotation pipelines ignore.

Methodologically, the study reflects the demanding craft of plasma proteomics. The authors employed filter-aided sample preparation, a technique that uses molecular-weight filters to clean and digest protein samples, and combined complementary fragmentation methods, collision-induced dissociation and higher-energy collisional dissociation, to maximize the information extracted from each peptide. Quantification was performed label-free, comparing spectral intensities across samples rather than relying on chemical tags introduced beforehand. The abbreviation NSRP1, for nuclear speckle splicing regulatory protein 1, appears in the study’s key terms, indicating that even proteins not normally thought of as plasma residents contributed identifiable acetylated peptides to the landscape.

The study design itself was deliberately controlled. Twenty patients with Alzheimer’s disease and twenty age-matched controls provide a matched comparison in which age, the strongest demographic risk factor, is held constant. The plasma samples were obtained from a commercial vendor operating under applicable regulations for the collection and handling of biological specimens, which the authors note meant separate ethics approval was not required. The work was funded by a Grant-in-Aid for Transformative Research Areas from Japan’s Ministry of Education, Culture, Sports, Science and Technology, with additional research funding from Abbott Laboratories, and one co-author is an employee of Abbott Japan, a disclosure the authors make explicitly in the paper.

What the study does not yet claim is a diagnostic test. A landscape of 350 acetylation sites across forty individuals is a foundation, not a finished clinical instrument. Validating any individual acetylation site as a biomarker will require larger and more diverse cohorts, longitudinal sampling, and confirmation that the modification patterns are reproducible across laboratories and platforms. The authors frame their contribution accordingly: as a resource that may contribute to a more extensive understanding of the biochemical properties of human plasma protein components in relation to Alzheimer’s disease pathogenesis. In a field racing toward blood-based early detection, the message of the study is that the blood itself still holds layers of information, written in chemical tags, that science is only now learning to read.

Subject of Research: Lysine acetylation profiling of plasma proteins in Alzheimer's disease using mass spectrometry

Article Title: Lysine acetylation-directed high-resolution plasma proteomic landscape of Alzheimer’s disease

Article References: Kozuka-Hata, H., Kitamura, A., Hiroki, T., Miyamura, N., Yoshimura, T., Tsumoto, K., & Oyama, M. (2026). Lysine acetylation-directed high-resolution plasma proteomic landscape of Alzheimer’s disease. Clinical Proteomics. https://doi.org/10.1186/s12014-026-09635-y

Image Credits: AI Generated

DOI: 10.1186/s12014-026-09635-y

Keywords: Alzheimer's disease, plasma proteomics, lysine acetylation, mass spectrometry, post-translational modification, apolipoprotein, short open reading frame, biomarkers, Clinical Proteomics, blood-based diagnostics, Lysine, acetylation-directed

The study, led by teams at University Medicine Greifswald in Germany and Utrecht University in the Netherlands, exploited the TIMES cohort, an acronym for Tracking Individuals Monthly for Evaluating Stability. Participants, 32 women and 19 men with a mean age of about 34 years, donated blood once a month for twelve months between June 2017 and May 2018. This design is unusual in plasma proteomics, where most biomarker studies still rely on cross-sectional snapshots of a single time point per subject. By sampling repeatedly, the researchers could separate the variation that occurs within a person over time from the variation that exists between people, a distinction that is fundamental to interpreting any clinical measurement.

What makes the study particularly persuasive is its dual-laboratory architecture. The Greifswald site and the Utrecht site processed and analyzed the same samples using deliberately different workflows. One laboratory used a bead-based SP3 digestion protocol, the other a conventional liquid-phase digest. One measured peptides on a timsTOF Pro2 mass spectrometer, the other on the newer timsTOF HT, which features a fourth-generation trapped ion mobility analyzer with fivefold greater ion charge capacity. Even the software pipelines diverged: Greifswald searched raw data with Spectronaut against a large database of more than 42,000 human protein entries, while Utrecht used DIA-NN with a compact plasma-specific database of just 2,445 entries built from the Human Protein Atlas. Quantification strategies differed too, with Greifswald favoring intensity-based absolute quantification and Utrecht using MaxLFQ normalization.

Despite this deliberate methodological divergence, the two datasets agreed remarkably well. After filtering out contaminants and the hypervariable regions of immunoglobulins, Greifswald quantified 274 proteins and Utrecht 188, with 163 proteins overlapping. Comparing the two laboratories sample by sample, the mean Spearman correlation coefficient across all samples was 0.908 and the mean Pearson correlation was 0.824, indicating strong agreement in both the ranking of proteins and their absolute abundances. To make the numbers clinically meaningful, both sites converted relative intensities into absolute concentrations using a panel of 37 housekeeping plasma proteins with well-established reference values. The resulting estimates landed squarely within known clinical ranges, a benchmark that mass spectrometry workflows have historically struggled to meet.

With the analytical foundation secured, the team turned to the central biological question: how stable is a person’s plasma proteome over a year, and how different are people from one another? The answer came vividly from an unsupervised dimensionality reduction technique called UMAP, which projects high-dimensional protein data into two dimensions while preserving neighborhood structure. When all monthly samples from all donors were projected, each person’s twelve samples clustered tightly together, while the clusters of different donors sat far apart. In other words, your plasma proteome in January looks far more like your own proteome in December than like anyone else’s at any time of year.

To quantify this, the researchers built a machine learning classifier, a support vector machine with a radial basis kernel, and challenged it to identify which donor a blinded sample came from. Across 1,000 iterations, the model achieved a median classification accuracy of 98 percent. The proteome, in effect, carries a personal signature as distinctive as a fingerprint. Notably, statistical adjustment for age, sex and body mass index did not significantly reduce the inter-donor variability, suggesting that these common demographic confounders do not drive the observed differences in this cohort. The clustering was so robust that it even exposed a suspected sample mix-up: the December sample labeled as donor 27 sat squarely within donor 4’s cluster, prompting the team to exclude it as a probable aliquoting or labeling error.

The most consequential finding concerns the immunoglobulins, the antibody family that constitutes roughly 20 percent of total plasma protein mass. Because antibodies are extraordinarily polymorphic, with variable regions numbering in the millions of possible sequences, most plasma proteomics studies either physically deplete them or quietly ignore them in data analysis. This study did the opposite, retaining and quantifying them using peptides unique to the constant regions of each antibody class and subclass. The measurements showed that immunoglobulins vary enormously between donors but barely at all within a donor over twelve months. Total IgG averaged 9.75 grams per liter across the cohort with a standard deviation of 2.64 grams per liter, a 27 percent coefficient of variation, while the average intra-donor fluctuation was only about 1.11 grams per liter. The same pattern held for IgG subclasses IgG1 through IgG4, as well as IgM, IgA1, IgA2 and IgD, and it was mirrored by CD5L, a protein known to travel in complex with IgM.

The mass spectrometry-derived subclass distribution also matched the literature closely. IgG1 averaged 6.12 milligrams per liter of blood plasma, representing about 62.8 percent of total IgG, compared with a literature value of roughly 63.6 percent. IgG2 came in at 19.1 percent versus an expected 24.5 percent, IgG3 at 9.3 percent versus 5.7 percent, and IgG4 at 8.7 percent versus 6.3 percent. These numbers matter because they demonstrate that untargeted mass spectrometry can now deliver antibody measurements of clinical grade without the targeted immunoassays traditionally required. They also align with recent work showing that the dominant antibody clones in plasma are produced by long-lived plasma cells, which would explain why circulating immunoglobulin levels remain so stable over months.

Not everything in plasma stood still. C-reactive protein, the classic inflammation marker, behaved in a completely different manner. Most donors kept CRP concentrations low throughout the year, but a few individuals showed dramatic spikes, with one donor ranging from a basal 0.16 nanomolar to a peak of 721 nanomolar, a roughly 4,500-fold swing. Others, including six donors, maintained persistently elevated basal CRP of around 50 nanomolar across all twelve months, hinting at low-grade chronic inflammation. The episodic peaks were mirrored by sister inflammation markers SAA1, SAA2, S100A8 and S100A9, strengthening the interpretation that they reflect acute inflammatory episodes such as minor infections. The contrast between rock-steady antibodies and volatile CRP illustrates a key principle for biomarker research: some proteins define who you are, while others report what is happening to you right now.

The study also delivered an unplanned clinical vignette. One donor, identified as donor 26, showed immunoglobulin levels four to ten times below the cohort average for IgG1, IgG2, IgG3, IgA1 and IgM, all falling below the 2.5th percentile, combined with high body mass index and chronically elevated CRP. This pattern is suggestive of common variable immunodeficiency, the most prevalent primary immunodeficiency in humans, although the donor reported no increased susceptibility to infections. The researchers emphasize that this chance observation, surfaced purely through longitudinal proteomic screening, illustrates the diagnostic potential of the approach. More broadly, the work argues that biomarker discovery needs longitudinal designs, donor-specific baselines and cross-laboratory validation, and that a universal plasma reference material or isotopically labeled internal standards should accompany multi-center studies. If a single blood draw can be read against a person’s own stable proteomic baseline rather than a population average, the sensitivity of future diagnostics could improve substantially.

Subject of Research: Longitudinal mass spectrometry analysis of inter-donor variability in the human plasma proteome, focusing on immunoglobulin subclass stability

Article Title: Immunoglobulin sub-class levels define inter-donor plasma variability: a longitudinal dual-lab study

Article References: Michalik, S., Kalaidopoulou Nteak, S., Drouin, N., Gesell Salazar, M., Hammer, E., Dhople, V. M., Holtfreter, S., Weiss, S., van den Toorn, H. W. P., Bröker, B. M., Domańska, G., Völker, U., & Heck, A. J. R. (2026). Immunoglobulin sub-class levels define inter-donor plasma variability: a longitudinal dual-lab study. Molecular Systems Biology, 22(8), 1312-1332. https://doi.org/10.1038/s44320-026-00218-5

Image Credits: AI Generated

DOI: 10.1038/s44320-026-00218-5

Keywords: plasma proteomics, immunoglobulins, mass spectrometry, longitudinal study, biomarkers, C-reactive protein, IgG subclasses, inter-donor variability, machine learning, clinical diagnostics, TIMES cohort, proteome stability

The research, led by Yeongshin Kim, Junho Park, Dongyoon Shin and Youngsoo Kim of CHA University in Seongnam, South Korea, set out to answer a deceptively simple question: do the elaborate plasma preparation platforms that have been benchmarked on cutting-edge mass spectrometers behave the same way when run on more accessible liquid chromatography-mass spectrometry, or LC-MS, instrumentation? The answer matters because most published evaluations of these platforms have relied on next-generation instruments with exceptional sensitivity and speed, leaving a genuine gap in knowledge for the many laboratories around the world that operate older or more modest equipment. If platform performance were fundamentally tied to top-tier hardware, the democratization of plasma biomarker discovery would stall; if not, the field could expand dramatically.

To address the question, the team processed a commercial pooled plasma sample through seven different preparation strategies. These included neat plasma with no treatment at all, serving as the baseline; the Multiple Affinity Removal System Human 14, known as MARS14, an antibody-based column that strips out the fourteen most abundant plasma proteins; and perchloric acid precipitation with neutralization, a chemical approach that preferentially precipitates abundant proteins while leaving many low-abundance species in solution. Alongside these depletion methods, the researchers tested four enrichment platforms: ENRICH-iST, Mag-Net, Proteonano and Proteograph XT. Each of these uses a different physical or chemical principle, from nanoparticle surfaces to specialized capture chemistries, to concentrate the dilute population of low-abundance proteins that standard workflows miss.

All samples were then analyzed in data-independent acquisition, or DIA, mode, a mass spectrometry strategy that systematically fragments all ions within defined mass windows rather than selecting individual precursors. DIA has become the workhorse of modern quantitative proteomics because it produces highly reproducible measurements across large sample cohorts, reducing the stochastic missing values that plague older data-dependent approaches. The choice of DIA was itself significant: it is the acquisition mode most commonly implemented on accessible instrumentation, so any conclusions drawn from the study would apply directly to the laboratories the researchers hoped to reach.

The results were striking. Untreated neat plasma yielded an average of just 777 identified proteins, a figure that captures the brutal reality of plasma’s dynamic range problem. The two depletion platforms performed substantially better, identifying between 1,278 and 1,468 proteins on average. But the enrichment platforms left the depletion approaches behind, with coverage ranging from 1,891 to a remarkable 6,060 proteins. Proteograph XT, the nanoparticle-based platform, delivered the deepest coverage of all and achieved the lowest missing value rate in the study, at just 1.9 percent, meaning that nearly every protein it detected was quantified consistently across the analysis rather than appearing and disappearing between runs.

Yet the study’s most important finding may be the one that complicates the simple narrative that more proteins equals better science. The researchers observed that greater proteome coverage did not always correlate with better quantitative precision. A platform that identifies thousands of additional proteins may do so at the cost of noisier measurements, and a coefficient of variation that looks acceptable for one platform may be unacceptable for another. Because the ultimate goal of plasma proteomics is reliable quantification, particularly when comparing patient cohorts to find proteins that distinguish disease from health, precision matters as much as depth. The study makes clear that platform choice significantly affected both the abundance profiles of the resulting datasets and the coverage of clinically relevant proteins, meaning that laboratories must match their preparation strategy to their biological question rather than simply chasing the largest protein count.

This platform-specific reshaping of the plasma proteome is a phenomenon that has been noted in previous work but never before systematically confirmed on accessible instrumentation. Each preparation method imposes its own bias: antibody-based depletion removes not only its targets but also proteins that travel bound to them, such as carrier proteins shuttling hormones and metabolites; chemical precipitation can lose proteins that co-precipitate with the abundant fraction; and nanoparticle enrichment selects for proteins with affinity for particular surface chemistries, so different particles fish out different, partially overlapping slices of the proteome. The consequence is that two laboratories studying the same plasma pool with different platforms may report substantially different protein abundance profiles, a fact that has major implications for reproducibility and for the design of multi-center biomarker studies.

Crucially, when the researchers compared their results with recent benchmarks generated on next-generation mass spectrometers, the platform-specific differences they observed reproduced those earlier findings. This is the study’s central vindication: the accessible LC-MS setups captured the same key distinctions among preparation platforms that the flagship instruments had revealed. In other words, the relative ranking of platforms, the patterns of proteome reshaping and the qualitative conclusions about which strategies best expose the low-abundance proteome all held true on standard equipment. What the newest instruments add is primarily depth and throughput, not a fundamentally different picture of how the platforms behave.

The practical implications extend well beyond methodological housekeeping. Plasma biomarker discovery has accelerated enormously in recent years, fueled by large-scale efforts such as the Human Plasma Proteome Project and by growing interest in early cancer detection, neurodegenerative disease monitoring and cardiovascular risk prediction. But much of that progress has been concentrated in well-funded centers equipped with the latest triple-TOF or Orbitrap Astral class instruments. If the platform evaluation framework demonstrated here holds, hospital-affiliated laboratories, research institutes in lower-resource settings and clinical translation teams can now participate meaningfully in plasma proteomics using instrumentation they already own. The study’s authors frame this as supporting the broader adoption of plasma proteomics in settings where next-generation instrumentation is not routinely available, and the data appear to justify that framing.

There remain caveats that the field will need to keep in view. The study used a single commercial pooled plasma sample, an excellent control for technical comparison but not a substitute for testing platforms across real patient cohorts with their biological variability, pre-analytical noise and disease-specific matrix effects. The evaluation also reflects one DIA acquisition strategy on one class of accessible instrument, and other configurations could shift the absolute numbers, even if the relative platform behavior is likely to persist. And the finding that coverage and precision can trade off against each other means that no single platform emerges as a universal winner; instead, the study provides something arguably more useful, a rigorous, reproducible map of what each of seven platforms delivers on hardware that most laboratories can access. As plasma proteomics moves from discovery science toward clinical application, that kind of practical, instrument-agnostic evidence may prove to be exactly what the field needs to turn an information-rich body fluid into a routine diagnostic resource.

Subject of Research: Comparative evaluation of plasma preparation platforms for proteomics using accessible liquid chromatography-mass spectrometry instrumentation

Article Title: Comprehensive evaluation of plasma proteomics platforms toward practical applications using accessible mass spectrometry instrumentation

Article References: Kim, Y., Park, J., Shin, D., & Kim, Y. (2026). Comprehensive evaluation of plasma proteomics platforms toward practical applications using accessible mass spectrometry instrumentation. Clinical Proteomics. https://doi.org/10.1186/s12014-026-09631-2

Image Credits: AI Generated

DOI: 10.1186/s12014-026-09631-2

Keywords: plasma proteomics, mass spectrometry, biomarker discovery, data-independent acquisition, protein depletion, nanoparticle enrichment, Proteograph XT, MARS14, LC-MS, proteome coverage, quantitative precision, Clinical Proteomics

The research, led by Zhenfang Shan, Jieqiong Li, and Zhaohui Tong of Beijing Chao-Yang Hospital, Capital Medical University, took a proteomics approach—measuring large panels of proteins in plasma rather than relying on clinical scores alone. The team analyzed plasma samples from 196 patients with ARDS, 100 pulmonary patients who did not have ARDS, and 100 healthy controls. Using mass spectrometry-based profiling and statistical tools including partial least squares-discriminant analysis, the researchers searched for proteins that consistently separated ARDS patients from the comparison groups.

The signal that emerged was striking. Pathways related to the proteasome—the cellular machine that shreds damaged or unwanted proteins into small peptides—were significantly activated in ARDS patients. Of particular interest was the immunoproteasome, a specialized version of the proteasome found in immune cells that processes proteins for presentation to the immune system. Rather than a single protein, the researchers found that a panel of seven immunoproteasome subunits collectively carried a robust diagnostic fingerprint of the syndrome.

To turn that observation into a practical tool, the team built a diagnostic model using LASSO regression, a machine learning technique that selects the most informative predictors while guarding against overfitting. Trained on 277 samples, the seven-subunit panel achieved an area under the receiver operating characteristic curve (AUC) of 0.908, indicating strong discriminatory power. Critically, the model held up in an independent validation cohort of 119 samples, where it achieved an AUC of 0.890. That kind of consistency across separate cohorts is what separates promising biomarkers from statistical artifacts, and it suggests the panel could distinguish ARDS from other serious lung conditions at the bedside.

The researchers did not stop at diagnosis. They traced where the biomarkers came from, examining bronchoalveolar lavage fluid—the liquid sampled from deep within the lungs of patients—and confirming that the proteasome signature reflected genuine biological activity in the injured lung rather than a distant systemic echo. They then moved into an animal model, inducing ARDS-like lung injury in mice with lipopolysaccharide, a component of bacterial cell walls that triggers intense inflammation.

In the mouse model, one subunit stood out: PSMB8, also known as the low molecular mass polypeptide 7 or LMP7, a catalytic subunit unique to the immunoproteasome. PSMB8 expression was upregulated in lung tissue after injury, and single-cell and lineage analyses indicated that the protein was produced primarily by monocytes and macrophages—the scavenger immune cells that swarm into inflamed lungs and often drive much of the collateral tissue damage. The degree of PSMB8 elevation correlated with disease severity, tying the molecular marker to clinical outcomes.

The therapeutic experiments were the most consequential part of the study. When the researchers inhibited PSMB8 pharmacologically, or deleted the Psmb8 gene in mice, the results were consistent: lung injury was reduced, inflammation was blunted, and respiratory dysfunction improved. Measurements of lung mechanics, including resistance and compliance, along with wet-to-dry weight ratios that quantify fluid accumulation in the lung, all pointed toward meaningful protection when the immunoproteasome subunit was removed from the picture.

The mechanism behind that protection appears to lie in macrophage behavior. PSMB8 deficiency suppressed M1-like macrophage polarization—the pro-inflammatory state that macrophages adopt when they are pumping out cytokines and reactive molecules that, while meant to fight pathogens, can devastate surrounding tissue. Intriguingly, the researchers found that this shift was accompanied by alterations in lipid metabolic pathways, hinting that the immunoproteasome influences not just protein degradation but also the metabolic programming that determines whether macrophages soothe or scorch the lung.

The implications are twofold. Diagnostically, a blood test built on seven proteasome subunits could complement existing tools such as the Sequential Organ Failure Assessment and APACHE II scores, offering an objective, biologically grounded way to identify ARDS early—potentially before patients require invasive mechanical ventilation. Therapeutically, PSMB8 joins a growing list of immune-focused targets for a syndrome that currently lacks a specific drug; treatment today is largely supportive, centered on ventilator management and care of the underlying cause.

Caveats remain. The therapeutic results come from a preclinical mouse model, and the authors themselves emphasize that PSMB8 is a promising but still preclinically validated target that warrants further investigation. Translating proteasome inhibition into safe human therapy will require careful work, since the immunoproteasome plays normal roles in immune surveillance. Still, the convergence of a clinically validated biomarker panel with a mechanistically supported drug target—grounded in more than 300 patient samples and confirmed in animal studies—represents an unusually coherent translational package for a syndrome that has long resisted both early detection and targeted treatment.

Subject of Research: A proteasome subunit-based biomarker panel for early diagnosis of acute respiratory distress syndrome and therapeutic targeting of PSMB8

Article Title: Proteasome subunit-based biomarker panel for early diagnosis of acute respiratory distress syndrome and therapeutic targeting of PSMB8 in a mouse model

Article References: Shan, Z., Zhang, H., Ma, J., Gou, Y., Li, J., & Tong, Z. (2026). Proteasome subunit-based biomarker panel for early diagnosis of acute respiratory distress syndrome and therapeutic targeting of PSMB8 in a mouse model. Journal of Translational Medicine. https://doi.org/10.1186/s12967-026-08645-6

Image Credits: AI Generated

DOI: 10.1186/s12967-026-08645-6

Keywords: ARDS, biomarkers, proteasome, immunoproteasome, PSMB8, macrophage polarization, plasma proteomics, LASSO regression, lung injury, intensive care, diagnostic panel, lipopolysaccharide mouse model

The logic behind the approach rests on a decade of progress in proteomics, the large-scale study of the protein complement of biological systems. Unlike the genome, which is essentially fixed at conception, the proteome is dynamic: it shifts with infection, stress, exercise, disease, and, crucially, with age itself. Researchers have shown that machine-learning models trained on thousands of plasma protein measurements can predict chronological age with remarkable accuracy, and that the difference between predicted and actual age—sometimes called proteomic age acceleration—correlates with frailty, multimorbidity, and mortality risk. These proteomic clocks are conceptually analogous to the epigenetic clocks based on DNA methylation that transformed aging research in the 2010s, but they offer distinct advantages for interventional trials: blood proteins are readily sampled repeatedly, they respond to physiological change on shorter timescales than DNA methylation marks, and many of the proteins involved are themselves drug targets or biomarkers already familiar to the pharmaceutical industry.

What the new study demonstrates is that this technology can survive contact with the realities of clinical trial design. The researchers integrated multiple proteomic aging clocks into the analytical framework of a phase 2a trial, the earliest stage at which a therapeutic candidate is tested in patients or at-risk volunteers to gather preliminary evidence of biological activity. Rather than treating biological age estimation as an exploratory afterthought, the team built the clocks into the statistical evaluation plan from the outset, defining in advance how changes in proteomic age would be measured, how measurement noise would be handled, and how multiple clock outputs could be combined to give a coherent picture of whether an intervention was shifting the biology of aging in a favorable direction.

The emphasis on simultaneous assessment is the study’s most distinctive contribution. Different aging clocks, trained on different protein sets and different cohorts, capture partly overlapping and partly distinct facets of the aging process—one model may be more sensitive to inflammatory pathways, another to metabolic or hepatic changes, a third to renal or cardiovascular decline. By deploying a suite of clocks in parallel rather than betting on a single algorithm, the trial design allows investigators to ask not merely whether an intervention changes a number, but whether it changes the underlying biology in a way that is consistent across independent measures of aging. Convergence across multiple clocks provides a form of internal replication that a single biomarker cannot offer, while divergence among clocks can itself be informative, pointing to organ-specific or pathway-specific effects that would otherwise be invisible.

Technically, the integration demanded solutions to several stubborn problems. Proteomic measurements are sensitive to pre-analytical variables: the choice of assay platform, the timing of blood draws, fasting status, and even the season of sample collection can shift protein concentrations. Longitudinal interpretation requires that the analytical pipeline distinguish true within-person change from batch effects and ordinary biological fluctuation. The study addressed these challenges by anchoring the clocks to repeated baseline sampling, applying rigorous quality control to the proteomic data, and using statistical models that estimate change within individuals rather than relying solely on comparisons between treatment and control groups at a single time point. This within-person framing is essential for short trials, because it dramatically increases statistical power when each participant serves as their own reference for the direction and magnitude of biological aging.

The broader significance of the work lies in what it could do to the economics of geroscience. Developing drugs that target aging has long been caught in a regulatory and commercial Catch-22: regulators generally approve treatments for diseases, not for aging itself, because aging lacks an agreed clinical endpoint; without approved indications, investment in geroprotective therapies has lagged. Biomarkers that can credibly demonstrate a slowing of biological aging over months rather than decades offer a path through this impasse. If proteomic clocks can be validated as surrogate endpoints or at least as robust pharmacodynamic markers, early-phase trials of candidate geroprotectors—whether repurposed drugs such as rapamycin and its analogues, senolytic agents that clear senescent cells, or novel molecules designed around aging pathways—become faster, smaller, and far cheaper to run.

The phase 2a setting is precisely where such markers earn their keep. Phase 2a studies are designed to detect signals of biological activity, not to prove clinical benefit, and a biomarker that reliably responds to an intervention’s mechanism of action is exactly the kind of signal these trials exist to find. Embedding proteomic clocks at this stage creates a screening funnel: interventions that show consistent effects across multiple aging measures can be advanced to larger trials with confidence, while those that leave the proteome untouched can be deprioritized before expensive late-stage development. In effect, the clocks function as a biological readout of geroprotection, analogous to how viral load measurements transformed the early development of antiretroviral therapies.

Important caveats remain, and the authors and the field are careful to acknowledge them. A change in a proteomic aging clock is not yet proof that a therapy extends healthspan or lifespan; the clocks are validated against age-related outcomes in observational data, and demonstrating that an intervention moves the biomarker is only the first step toward showing that it changes disease trajectories. Calibration across diverse populations is another open question, since proteomic aging signatures can vary with ancestry, sex, socioeconomic factors, and baseline health status, and a clock optimized in one cohort may miscalibrate in another. Standardization across assay platforms and laboratories will also be necessary before proteomic age becomes a measure that regulators and clinicians can compare across studies. The new work does not resolve these issues single-handedly, but it establishes a concrete, tested framework within which they can be addressed trial by trial.

Even so, the moment feels like an inflection point. The geroscience field has spent years generating compelling animal data on interventions that delay aging, only to face a translational bottleneck at the human frontier. The integration of proteomic aging clocks into a real phase 2a clinical trial shows that the measurement problem—long the field’s most fundamental limitation—is tractable with current technology. If subsequent trials replicate and extend this framework, the result could be a virtuous cycle in which better biomarkers enable faster trials, faster trials attract greater investment, and greater investment produces the interventions that finally move the needle on human healthspan. Aging, for the first time, is being measured in the clinic on the timescale of a clinical trial, and that change may prove as important as any single therapeutic candidate now in development.

Subject of Research: Proteomic aging clocks integrated into a phase 2a clinical trial for geroprotective assessment

Article Title: Integration of proteomic aging clocks in a phase 2a clinical trial supports simultaneous geroprotective assessment

Article References: Zhavoronkov, A., Galkin, F., Chen, S., Ren, F., Aliper, A., Durymanov, M., Sidorenko, D., Cui, H., Han, J.-D. J., Xu, H., Liu, X., Xu, Z., Kuppe, C., Austin Argentieri, M., Ying, K., Goeminne, L. J. E., Moqri, M., Tyshkovskiy, A., & Gladyshev, V. N. (2026). Integration of proteomic aging clocks in a phase 2a clinical trial supports simultaneous geroprotective assessment. Nature Biotechnology. https://doi.org/10.1038/s41587-026-03286-y

Image Credits: AI Generated

DOI: 10.1038/s41587-026-03286-y

Keywords: proteomic aging clocks, biological age, geroprotection, phase 2a trial, geroscience, biomarkers, plasma proteomics, aging interventions, healthspan, clinical trial design, machine learning, surrogate endpoints

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