Biology subjects

Apostolidis, S. A.

Publications and source records attributed to Apostolidis, S. A..

9 recordsLinked to original sources

bioRxiv2026

Limits of marker information transfer]{Limits of marker information transfer in cross-panel cytometry imputation

Marker imputation can bridge cytometry datasets acquired with different panels and platforms. Its validity depends on whether common markers retain information about unmeasured target markers and whether those relationships transfer across assay contexts. We evaluated these limits in two paired flow cytometry-CyTOF cohorts comprising 134 patients and 257 specimens. A patient-held-out framework distinguished within-platform same-cell reconstruction from cross-platform distribution translation without cell correspondence. We compared conventional machine-learning and deep generative estimators with median imputation and used permutation controls to estimate the contributions of patient matching and source-exclusive markers. Evaluation covered marker values, specimen distributions, phenotype utility, and directional predictive relationships among markers. Within platforms, exact k-nearest neighbors reduced mean absolute error by 12.8-35.0% relative to median imputation. However, the median-based reconstructions were contracted, with a median ratio of predicted to observed interquartile range below 1 in every context and evaluable population. Patient matching improved cross-platform distribution recovery, whereas adding source-exclusive markers or higher-capacity estimators did not consistently produce net gains. In the independently annotated acute myeloid leukemia (AML) cohort, reconstructed markers largely preserved broad-lineage classification but did not recover fine-phenotype information available through direct measurement. In AML-CyTOF, the measured full panel outperformed the reconstructed representation by 10.25 percentage points for fine phenotypes. Predictive relationships among markers were context-specific. Most targets were better explained by several markers than by one. Marker selection based on these relationships showed a trade-off between panel size and pairwise coverage. Paired samples and imputation can therefore connect selected marker distributions across panels but should not be treated as substitutes for direct measurement. Cross-panel cytometry should be evaluated with patient-held-out tests matched to the resolution of available ground truth and the intended downstream use.

bioinformatics↗

bioRxiv2026

CytoGate-Bench: an LLM benchmark for cross-panel cell gating in cytometry

In cytometry, the workhorse single-cell technology of clinical immunology, every study defines its own antibody panel and cell-type vocabulary, so a classifier trained on one cannot annotate the next. Immunologists instead annotate by manual gating, splitting one parent population at a time on a two-marker plot, down an expert-defined hierarchy. We introduce CytoGate-Bench, a benchmark that reformulates this per-step procedure as a zero-shot, panel-agnostic task for large language models. It comprises 23,646 expert-annotated instances re-curated from 11 public flow- and mass-cytometry cohorts spanning eight marker panels. Across six open- and closed-weight backbones, the strongest formulation draws one rectangular gate per candidate and falls within the range of trained, panel-specialized baselines. It degrades less under distribution shift. Walking the hierarchy stepwise outperforms predicting every cell type at once. Ablations trace the signal to the data distribution shape and curated marker priors. However, adding vision or a self-verification loop systematically tightens gates.

bioinformatics↗

bioRxiv2026

Deep learning representations of human Immune Health for precision immunology

The human immune system is composed of [~]30-50 distinct cell types, each of which can exist in different states of activation or differentiation. Indeed, the mammalian immune system has evolved to sense and respond to infections, cancers, injuries, and changes in tissue or host homeostasis (1). Moreover, an increasingly large fraction of approved drugs target the immune system directly, and/or cause immune changes (2-4). A key feature of the immune system is to store some of this information, for example as innate or adaptive immune memory (5). In addition, rewiring of immune network architecture induced by disease, environmental exposures, drug treatments, and/or chronological age allows the immune system to store information in the pattern of connections and activity across populations of immune cells. This ensemble information storage, in addition to changes to individual cells, functions as a major way the immune system encodes aspects of immune history and future potential. Genetic information can identify inherited risk alleles, but cannot capture the continual remodeling of the immune system shaped by exposures, infection, inflammation, therapy, and aging (6, 7). To define and use such ensemble immunotypes, we developed a self-supervised deep learning framework that transforms high-dimensional immune profiles into representations of immune health. MAESTRO (MAsked Encoding Set TRansformer with self-distillatiOn) encodes a set of cells from an individual into an embedding that captures immune cell population-level organization. Pretrained on 1,792 peripheral blood samples comprising over 418 million immune cells across 13 clinical diagnoses, MAESTRO learns immune fingerprints that are stable within individuals yet diverse across populations, states of health, disease, and treatment, providing a quantitative basis for comparing immune states across individuals and over time. These fingerprints capture immune architecture beyond coarse cell type proportions, enabling patient-efficient clinical prediction using simple task specific models. MAESTRO model embeddings retain a temporal dimension of immune history and potential, reflecting signatures of past exposures and baseline features that predict future immune responses. Finally, we demonstrate a translational precision immunotherapy application by testing this approach in metastatic Pancreatic Ductal Adenocarcinoma (PDAC), where pretreatment immune landscape circuitry maps enable patient stratification and therapeutic response prediction. Overall, we developed a large, attention-based model that captures deep network architecture of immune states through self- supervised representations of immune cytometry data as a reusable foundation for precision immunology, converting immune complexity into clinically actionable embeddings for diagnosis, monitoring, and therapy selection.

immunology↗

bioRxiv2024

Single-cell Masked Autoencoder: An Accurate and Interpretable Automated Immunophenotyper

High-throughput single-cell cytometry data are crucial for understanding involvement of immune system in diseases and responses to treatment. Traditional methods for annotating cytometry data, specifically manual gating and clustering, face challenges in scalability, robustness, and accuracy. In this study, we propose a cytometry masked autoencoder (cyMAE), which offers an automated solution for immunophenotyping tasks including cell type annotation. The cyMAE model is designed to uphold user-defined cell type definitions, thereby facilitating easier interpretation and cross-study comparisons. The cyMAE model operates on a pre-train and fine-tune approach. In the pre-training phase, cyMAE employs Masked Cytometry Modelling (MCM) to learn relationships between protein markers in immune cells solely based on protein expression, without relying on prior information such as cell identity and cell type-specific marker proteins. Subsequently, the pre-trained cyMAE is fine-tuned on multiple specialized tasks via task-specific supervised learning. The pre-trained cyMAE addresses the shortcomings of manual gating and clustering methods by providing accurate and interpretable predictions. Through validation across multiple cohorts, we demonstrate that cyMAE effectively identifies co-occurrence patterns of bound labeled antibodies, delivers accurate and interpretable cellular immunophenotyping, and improves the prediction of subject metadata status. Specifically, we evaluated cyMAE for cell type annotation and imputation at the cellular-level and SARS-CoV-2 infection prediction, secondary immune response prediction against COVID-19, and prediction of the infection stage in COVID-19 progression at the subject-level. The introduction of cyMAE marks a significant step forward in immunology research, particularly in large-scale and high-throughput human immune profiling. This approach offers new possibilities for predicting and interpreting cellular-level and subject-level phenotypes in both health and disease.

bioinformatics↗

bioRxiv2023

Prior vaccination enhances immune responses during SARS-CoV-2 breakthrough infection with early activation of memory T cells followed by production of potent neutralizing antibodies

SARS-CoV-2 infection of vaccinated individuals is increasingly common but rarely results in severe disease, likely due to the enhanced potency and accelerated kinetics of memory immune responses. However, there have been few opportunities to rigorously study early recall responses during human viral infection. To better understand human immune memory and identify potential mediators of lasting vaccine efficacy, we used high-dimensional flow cytometry and SARS-CoV-2 antigen probes to examine immune responses in longitudinal samples from vaccinated individuals infected during the Omicron wave. These studies revealed heightened Spike-specific responses during infection of vaccinated compared to unvaccinated individuals. Spike-specific CD4 T cells and plasmablasts expanded and CD8 T cells were robustly activated during the first week. In contrast, memory B cell activation, neutralizing antibody production, and primary responses to non-Spike antigens occurred during the second week. Collectively, these data demonstrate the functionality of vaccine-primed immune memory and highlight memory T cells as rapid responders during SARS-CoV-2 infection.

immunology↗

bioRxiv2022

Efficient recall of Omicron-reactive B cell memory after a third dose of SARS-CoV-2 mRNA vaccine

Despite a clear role in protective immunity, the durability and quality of antibody and memory B cell responses induced by mRNA vaccination, particularly by a 3rd dose of vaccine, remains unclear. Here, we examined antibody and memory B cell responses in a cohort of individuals sampled longitudinally for [~]9-10 months after the primary 2-dose mRNA vaccine series, as well as for [~]3 months after a 3rd mRNA vaccine dose. Notably, antibody decay slowed significantly between 6- and 9-months post-primary vaccination, essentially stabilizing at the time of the 3rd dose. Antibody quality also continued to improve for at least 9 months after primary 2-dose vaccination. Spike- and RBD-specific memory B cells were stable through 9 months post-vaccination with no evidence of decline over time, and [~]40-50% of RBD-specific memory B cells were capable of simultaneously recognizing the Alpha, Beta, Delta, and Omicron variants. Omicron-binding memory B cells induced by the first 2 doses of mRNA vaccine were boosted significantly by a 3rd dose and the magnitude of this boosting was similar to memory B cells specific for other variants. Pre-3rd dose memory B cell frequencies correlated with the increase in neutralizing antibody titers after the 3rd dose. In contrast, pre-3rd dose antibody titers inversely correlated with the fold-change of antibody boosting, suggesting that high levels of circulating antibodies may limit reactivation of immunological memory and constrain further antibody boosting by mRNA vaccines. These data provide a deeper understanding of how the quantity and quality of antibody and memory B cell responses change over time and number of antigen exposures. These data also provide insight into potential immune dynamics following recall responses to additional vaccine doses or post-vaccination infections. Graphical Summary O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=123 SRC="FIGDIR/small/481163v1_ufig1.gif" ALT="Figure 1"> View larger version (20K): org.highwire.dtl.DTLVardef@123d2d9org.highwire.dtl.DTLVardef@e7db82org.highwire.dtl.DTLVardef@1fc73deorg.highwire.dtl.DTLVardef@11b21f9_HPS_FORMAT_FIGEXP M_FIG C_FIG

immunology↗

bioRxiv2021

mRNA Vaccination Induces Durable Immune Memory to SARS-CoV-2 with Continued Evolution to Variants of Concern

SARS-CoV-2 mRNA vaccines have shown remarkable efficacy, especially in preventing severe illness and hospitalization. However, the emergence of several variants of concern and reports of declining antibody levels have raised uncertainty about the durability of immune memory following vaccination. In this study, we longitudinally profiled both antibody and cellular immune responses in SARS-CoV-2 naive and recovered individuals from pre-vaccine baseline to 6 months post-mRNA vaccination. Antibody and neutralizing titers decayed from peak levels but remained detectable in all subjects at 6 months post-vaccination. Functional memory B cell responses, including those specific for the receptor binding domain (RBD) of the Alpha (B.1.1.7), Beta (B.1.351), and Delta (B.1.617.2) variants, were also efficiently generated by mRNA vaccination and continued to increase in frequency between 3 and 6 months post-vaccination. Notably, most memory B cells induced by mRNA vaccines were capable of cross-binding variants of concern, and B cell receptor sequencing revealed significantly more hypermutation in these RBD variant-binding clones compared to clones that exclusively bound wild-type RBD. Moreover, the percent of variant cross-binding memory B cells was higher in vaccinees than individuals who recovered from mild COVID-19. mRNA vaccination also generated antigen-specific CD8+ T cells and durable memory CD4+ T cells in most individuals, with early CD4+ T cell responses correlating with humoral immunity at later timepoints. These findings demonstrate robust, multi-component humoral and cellular immune memory to SARS-CoV-2 and current variants of concern for at least 6 months after mRNA vaccination. Finally, we observed that boosting of pre-existing immunity with mRNA vaccination in SARS-CoV-2 recovered individuals primarily increased antibody responses in the short-term without significantly altering antibody decay rates or long-term B and T cell memory. Together, this study provides insights into the generation and evolution of vaccine-induced immunity to SARS-CoV-2, including variants of concern, and has implications for future booster strategies. GRAPHICAL ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=146 HEIGHT=200 SRC="FIGDIR/small/457229v1_ufig1.gif" ALT="Figure 1"> View larger version (32K): org.highwire.dtl.DTLVardef@16c64b1org.highwire.dtl.DTLVardef@146ca3aorg.highwire.dtl.DTLVardef@86b7edorg.highwire.dtl.DTLVardef@956879_HPS_FORMAT_FIGEXP M_FIG C_FIG

immunology↗

bioRxiv2021

Signaling through FcγRIIA and the C5a-C5aR pathway mediates platelet hyperactivation in COVID-19

Patients with COVID-19 present with a wide variety of clinical manifestations. Thromboembolic events constitute a significant cause of morbidity and mortality in patients infected with SARS-CoV-2. Severe COVID-19 has been associated with hyperinflammation and pre-existing cardiovascular disease. Platelets are important mediators and sensors of inflammation and are directly affected by cardiovascular stressors. In this report, we found that platelets from severely ill, hospitalized COVID-19 patients exhibit higher basal levels of activation measured by P-selectin surface expression, and have a poor functional reserve upon in vitro stimulation. Correlating clinical features to the ability of plasma from COVID-19 patients to stimulate control platelets identified ferritin as a pivotal clinical marker associated with platelet hyperactivation. The COVID-19 plasma-mediated effect on control platelets was highest for patients that subsequently developed inpatient thrombotic events. Proteomic analysis of plasma from COVID-19 patients identified key mediators of inflammation and cardiovascular disease that positively correlated with in vitro platelet activation. Mechanistically, blocking the signaling of the Fc{gamma}RIIa-Syk and C5a-C5aR pathways on platelets, using antibody-mediated neutralization, IgG depletion or the Syk inhibitor fostamatinib, reversed this hyperactivity driven by COVID-19 plasma and prevented platelet aggregation in endothelial microfluidic chamber conditions, thus identifying these potentially actionable pathways as central for platelet activation and/or vascular complications in COVID-19 patients. In conclusion, we reveal a key role of platelet-mediated immunothrombosis in COVID-19 and identify distinct, clinically relevant, targetable signaling pathways that mediate this effect. These studies have implications for the role of platelet hyperactivation in complications associated with SARS-CoV-2 infection. Cover illustration O_FIG_DISPLAY_L [Figure 1] M_FIG_DISPLAY C_FIG_DISPLAY One-sentence summaryThe Fc{gamma}RIIA and C5a-C5aR pathways mediate platelet hyperactivation in COVID-19

immunology↗

bioRxiv2021

Rapid induction of antigen-specific CD4+ T cells guides coordinated humoral and cellular immune responses to SARS-CoV-2 mRNA vaccination

The SARS-CoV-2 mRNA vaccines have shown remarkable clinical efficacy, but questions remain about the nature and kinetics of T cell priming. We performed longitudinal antigen-specific T cell analyses in healthy individuals following mRNA vaccination. Vaccination induced rapid near-maximal antigen-specific CD4+ T cell responses in all subjects after the first vaccine dose. CD8+ T cell responses developed gradually after the first and second dose and were variable. Vaccine-induced T cells had central memory characteristics and included both Tfh and Th1 subsets, similar to natural infection. Th1 and Tfh responses following the first dose predicted post-boost CD8+ T cell and neutralizing antibody levels, respectively. Integrated analysis of 26 antigen-specific T cell and humoral responses revealed coordinated features of the immune response to vaccination. Lastly, whereas booster vaccination improved CD4+ and CD8+ T cell responses in SARS-CoV-2 naive subjects, the second vaccine dose had little effect on T cell responses in SARS-CoV-2 recovered individuals. Thus, longitudinal analysis revealed robust T cell responses to mRNA vaccination and highlighted early induction of antigen-specific CD4+ T cells. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=192 HEIGHT=200 SRC="FIGDIR/small/440862v1_ufig1.gif" ALT="Figure 1"> View larger version (52K): org.highwire.dtl.DTLVardef@db812dorg.highwire.dtl.DTLVardef@fdc549org.highwire.dtl.DTLVardef@a34663org.highwire.dtl.DTLVardef@1621e52_HPS_FORMAT_FIGEXP M_FIG C_FIG

immunology↗