Predictive Signatures Research Articles

On double monsoon onset and its predictive signature

On double monsoon onset and its predictive signature

Identification of Interleukin-Related Genes Signature for Prognosis Prediction in Head and Neck Squamous Cell Carcinoma Patients.

This study focused on identifying the interleukin (IL)-related genes that influence the head and neck squamous cell carcinoma (HNSCC) patients' prognosis and response to anticancer therapy in patients with HNSCC. We developed a risk model that included three gene signatures, IL Enhancer Binding Factor 2 (ILF2), IL 36 alpha (IL36A), and IL10, based on differential expression analysis, survival analysis, Least Absolute Shrinkage and Selection Operator (LASSO) analysis, and Cox regression analysis. We found that the low-risk group was scored with higher immune cell infiltration, higher expression of human leukocyte antigen (HLA) family genes and immune checkpoint genes, higher cytolytic activity (CYT), tertiary lymphoid structures (TLS), and CD8A/PD-L1 ratio. In contrast, the high-risk group was scored with higher tumor immune dysfunction and exclusion (TIDE), which implied worse response to immunotherapy and worse prognosis. The results above indicated that the low-risk group had stronger antitumor immunity and better responsiveness to immunotherapy. We also observed a significantly enriched pattern of cancer-related pathways and immune pathways in the comparison of the high-risk and low-risk groups. Furthermore, the high-risk group had higher sensitivity to chemotherapy drugs, which suggested that they might benefit from chemotherapy treatment. Following the results above, we confirmed in HNSCC cell lines and clinical specimens that the level of ILF2 in tumors was significantly higher than that in adjacent tumor tissues. Besides, in vivo and in vitro results both showed that silencing ILF2 might depress tumor growth, invasion, and migration. This study not only provided novel perspectives into the immunological and molecular mechanisms of HNSCC and uncovered IL-related gene signatures for predicting HNSCC patients' prognosis and response to chemotherapy and immunotherapy, but also preliminarily suggested that ILF2 might be an important target in the treatment of HNSCC.

GULP1 as a Novel Diagnostic and Predictive Biomarker in Hepatocellular Carcinoma.

Hepatocellular carcinoma (HCC) is characterized by high recurrence and mortality, necessitating the identification of reliable biomarkers. In this study, we aimed to identify the predictive gene signatures for HCC recurrence and evaluate the efficiency of GULP PTB domain-containing engulfment adaptor 1 (GULP1) as a predictive and diagnostic marker and therapeutic target for HCC. We analyzed genomic datasets from The Cancer Genome Atlas and Gene Expression Omnibus databases via least absolute shrinkage and selection operator Cox regression and 10-fold cross-validation, leading to the development of a 15-gene risk score model, which was validated using three independent datasets. Serum GULP1 and α-fetoprotein levels were assessed to determine the diagnostic accuracy of the model. Using clinical cohorts and patient sera, GULP1 roles were examined, and functional assays in vitro and in vivo were used to evaluate its effects on cell growth, epithelial-mesenchymal transition (EMT), ADP-ribosylation factor 6 activation, and β-catenin signaling. Our newly developed risk-score model accurately predicted recurrent HCC in all datasets. Among the 15 genes in the risk score model, GULP1 was overexpressed in patients with HCC and independently predicted HCC recurrence. Its expression modulation influenced cell growth and EMT, with observed effects on ADP-ribosylation factor 6 activation and β-catenin signaling pathways. GULP1 is a crucial biomarker for HCC, serving as a non-invasive diagnostic and predictive tool. It also plays key roles in HCC progression. Our findings highlight the potential use of GULP1 in treatment strategies targeting EMT and HCC recurrence to improve the personalized care and patient outcomes.

Machine learning–based liquid biomarker analysis from the NEONAX trial for response prediction to perioperative (PO) and adjuvant (A) gemcitabine/nab-paclitaxel in resectable PDAC (rPDAC).

755 Background: The phase II NEONAX trial examined PO and A gemcitabine/nab-paclitaxel (G/nP) efficacy for rPDAC pts with a comprehensive biomarker program. This translational study aimed at identifying liquid biomarker for early prediction of G/nP success in each treatment group (PO and A) and for the combined group using machine learning (ML). Methods: Pts from both study arms were stratified into groups based on Short- or Long-DFS ( n=80 pts total; PO n=42 pts; A n=38 pts). Blood plasma from selected G/nP naïve pts was analyzed by multiplexed ELISA (mELISA, 80-Plex ProcartaPlex Human Immune Response) and mass spectrometry (MS; feature list generation) and clinical data were acquired for each patient. For ML pts were divided into training (80%) and validation (20%) datasets. The Weka-based algorithm WrapperSubsetEval (WSE) with 8 different classifiers was used for feature selection. The best performing (highest accuracy, best ROC-AUC, minimal signature) classifier with the corresponding feature panel was selected by a 10x10-fold cross-validation (CV). Respective panels for all features combined as compared to clinical features and Ca19-9 blood levels were tested for performance via CV and bootstrap aggregating for training and validation datasets (10x with replacement). Results: The feature generation process generated 579 features from G/nP-naive pts (mELISA: 80; clinical data: 99; MS: 400). For response prediction of the whole rPDAC group to G/nP (S-/L-DFS), the ML-based flow identified a panel of 8 proteins from MS (SERPINA1, C1QB, KRT1, C4B, UBB, VCAM1, HPD, LYZ) and 3 proteins from mELISA (Galectin3, IL34, CCL4) combined with a logistic regression classifier. The predictive panel with the best performance for determining PO G/nP success was established using a kernel logistic regression classifier and included 2 proteins from mELISA (CXCL2, IL17A), 4 proteins from MS (C3, IGLV3-25, HLA-B, SHBG), and 2 clinical data (WHO grade, Na/Mg blood level ratio). The best performing biomarker for predicting success in A included 2 clinical data (tumor size at staging, hematocrit) and 1 protein from mELISA (IL22) and was established with a random forest classifier. All panels represented minimal feature signatures (rPDAC: 11; PO: 8; A; 3) and showed a high performance with ROC-AUC (training) > 0.90, ROC-AUC (CV) > 0.85, and ROC-AUC (validation) > 0.90. All panels described were superior compared to similar predictive signatures (with corresponding ML classifiers) for all clinical data or Ca19-9 blood levels. Conclusions: We show that minimal liquid biomarker signatures for early prediction of G/nP success in the NEONAX trial based on mELISA, MS and clinical data can be established by ML. The study shows the potential of ML for biomarker panel development and the value of mELISA/MS for generation of feature lists to use in ML. Clinical trial information: NCT02047513 .

Predictive methylation signature for gemcitabine sensitivity in pancreatic ductal adenocarcinoma: A pathway to personalized treatment.

768 Background: Gemcitabine (Gem) is a key drug used in chemotherapy regimens for treating pancreatic ductal adenocarcinoma (PDA). Identifying a biomarker to predict patient response to Gem could be crucial in personalizing treatment selection between Gemcitabine and fluorouracil (5-FU)-based therapies, thereby improving patient outcomes. We hypothesize that DNA methylation changes in genes linked to Gem resistance may serve as surrogates for the expression of corresponding proteins, thereby predicting treatment outcomes in patients with PDA receiving this therapy. Methods: We curated a 93-gene panel through a literature review spanning from January 1970 to April 2024. The proteins encoded by these genes have demonstrated either preclinical evidence (from cell line or mouse models) or clinical evidence (from tissue-based studies) of affecting the response to Gem in PDA. We evaluated the prognostic significance of methylation at specific CpG (cytosine-phosphate-guanine) sites within these genes using data from The Cancer Genome Atlas (TCGA) database, focusing on patients who had received Gem (106 /184 patients). We utilized elastic net multivariate analysis to analyze survival and created Kaplan-Meier plots based on our gene panel to estimate overall survival (OS). We compared RNA and gene expression levels for the clinically significant gene signatures between normal (NT) and PDA tissues using TNMplot.com, a web-based tool developed from data in the Gene Expression Omnibus of the National Center for Biotechnology Information (NCBI-GEO), TCGA, Therapeutically Applicable Research to Generate Effective Treatments (TARGET), and The Genotype-Tissue Expression (GTEx) repositories. This tool employs Mann-Whitney or Kruskal-Wallis tests to determine statistical significance. Results: We identified an 8-CpG site methylation signature across eight genes, distinguishing a high-risk subgroup within the Gem-treated cohort in the TCGA database. Patients exhibiting methylation at specific CpG sites in this signature experienced significantly worse OS, with median survival times of 20.35 months compared to 49.38 months (p=0.0068). Additionally, the RNA expression (fold change: 2.11, p=2.12e-55) and gene expression (fold change: 1.42, p=3.14e-18) of these genes were markedly higher in PDA tissues compared to normal tissues. Notably, the candidate genes in our signature do not overlap with those used to classify PDA into classical versus basal subtypes. In an ongoing study, we are also investigating methylation changes of these candidate genes in the blood (cell-free DNA) of PDA patients. Conclusions: We present a signature that can predict Gem sensitivity or resistance in patients with PDA. This signature serves as a critical first step toward personalizing treatment strategies.

A Novel Platinum-Resistance-related Gene Signature in Ovarian Cancer: Identification and Patient-derived Organoids Verification.

Platinum-based chemotherapy resistance is one of the main contributors to the mortality of Ovarian Cancer (OC). It is believed that sensitive biomarkers for identifying the population that is platinum-resistant are urgently needed. This study aims to develop a platinum-resistance gene-based signature to predict OC patients' responses to platinum drugs as well as survival outcomes. A platinum-resistance-related gene model was built by bioinformatics analysis. Then, its predictive power was internally validated. Continually, a nomogram was constructed to confirm the model's predictive ability. Afterward, GSEA was used to explore our model's potential functions. The ESTIMATE, CIBERSORT, TIMER, and ssGSEA were applied to estimate immune conditions. Then, somatic mutation and drug sensitivity were also analyzed. Finally, to gain insights into the roles of targeted genes in drug sensitivity, patient-derived tumor organoids (PDOs) validation was performed. Nine platinum-resistance-related genes, including SLC22A2, TAP1, PC, MCM3, GTF2H2, FXYD5, SUPT6H, IGKC, and MATN2, were anchored to build the predictive model, which was well internally validated. Subsequently, GSEA unveiled that our model genes enriched in the Hedgehog signaling pathway. The predictive signature was associated with immune checkpoint inhibitors such as PD-1, PD-L1, and CTLA4, guiding immunotherapy applications for OC patients. Drugs such as dasatinib, midostaurin, metformin, MK-2206, and mitomycin C might also benefit OC patients with different risk scores. PDOs showed patients with high-risk scores were more resistant to cisplatin than patients with low-risk scores. The platinum-resistance-related gene signature (SLC22A2, TAP1, PC, MCM3, GTF2H2, FXYD5, SUPT6H, IGKC, and MATN2) is valuable for prognosis prediction and guidance of treatment choices for OC patients.

P-692. Early host blood transcriptional signatures predictive of lower respiratory tract disease progression during parainfluenza infection in hematopoietic cell transplant (HCT) recipients

Abstract Background Parainfluenza virus (PIV) lower respiratory tract infection (LRTI) following HCT is associated with increased mortality and post-transplant complications. Clinical risk scores for progression from upper respiratory tract infection (URTI) to LRTI have limited predictive value and host transcriptional profiles may improve risk stratification.Figure 1.URTI samples from proven LRTI progressors showed robust differences in transcriptional signatures compared to URTI samples from non-progressors. Significant genes identified from comparisons of URTI samples from A) proven LRTI progressors vs non-progressors and B) possible LRTI vs non-progressors. Volcano plots depict significantly overexpressed (blue) and underexpressed (red) genes. Y-axis denotes the p-values while x-axis denotes estimates for each comparison. Horizontal dashed line denotes an adjusted p-value cut off < 0.1. Methods All blood samples were collected in HCT recipients at the time of PIV URTI. Subjects were followed for progression to LRTI and classified as: URTI (non-progressors), possible LRTI progressors (upper tract PIV detection with radiographic changes), or proven LRTI progressors (lower tract PIV detection with radiographic changes). Libraries were prepared using TruSeq mRNA Stranded Library Kit and sequenced on the NovaSeq 6000. Linear models of gene expression and comparisons for “possible LRTI vs URTI” and “proven LRTI vs URTI“ were fit using kimma. An adjusted p-value < 0.1 (Benjamini-Hochberg) defined significant genes. Gene ontology enrichment analyses of overexpressed and underexpressed genes were performed using clusterProfiler.Figure 2.Gene ontology (GO) enrichment analyses of overexpressed (OE) and underexpressed (UE) genes in URTI samples from proven LRTI progressors.A) Dotplot of top GO terms within the Biological Processes (BP) sub-ontology enriched in each OE and UE gene cluster from proven-LRTI vs URTI comparisons. B) Treemaps of reduced GO terms underexpressed in URTI samples from proven LRTI progressors. Terms are grouped (colored) based on their reduced parent terms (white fonts) and the space used by the term is proportional to the enrichment score. Results 46 subjects with PIV infection following HCT (2010 - 2019) were included (URTI = 32; possible LRTI = 7; proven LRTI = 7). Robust differences in gene expression (881 genes, Fig. 1A) were observed in URTI samples from proven progressors relative to non-progressors, while possible LRTI progressors showed higher similarity to non-progressors (102 genes, Fig. 1B). Notably, processes involved in rRNA metabolism, ribosome biogenesis, and T cell differentiation were underexpressed in proven LRTI progressors, whereas immunoinflammatory pathways were overexpressed (Fig. 2A and 2B). Conclusion We demonstrated that early PIV URTI blood samples can identify predictive signatures of progression to LRTI in HCT recipients presenting with URTI. Underexpression of T cell immunity and ribosome/rRNA metabolic pathways in proven progressors suggest a protective role in reducing risk of progression to PIV LRTI in HCT recipients. Early stratification of patients at risk of progression using early host blood transcriptional signatures may be useful to guide treatment management, reducing overall mortality rate and disease burden in HCT recipients. Disclosures Chikara Ogimi, MD, PhD, AstraZeneca: Honoraria|bioMerieux Japan Ltd.: Honoraria|ELSEVIER: Honoraria|KYORIN: Honoraria|Miyarisan: Honoraria|MSD: Honoraria|NOVARTIS: Honoraria|Pfizer: Honoraria Michael J. Boeckh, MD PhD, Allovir: Advisor/Consultant|Allovir: Grant/Research Support|AstraZeneca: Advisor/Consultant|AstraZeneca: Grant/Research Support|Merck: Advisor/Consultant|Merck: Grant/Research Support|Moderna: Advisor/Consultant|Moderna: Grant/Research Support|Symbio: Advisor/Consultant Alpana waghmare, MD, Allovir: Grant/Research Support|Ansun Biopharma: Grant/Research Support|GlaxoKlineSmith: Advisor/Consultant|GlaxoKlineSmith: Grant/Research Support|Pfizer: Grant/Research Support|Vir: Advisor/Consultant

Robust Cluster Prediction Across Data Types Validates Association of Sex and Therapy Response in GBM

Background: Previous studies have described sex-specific patient subtyping in glioblastoma. The cluster labels associated with these “legacy data” were used to train a predictive model capable of recapitulating this clustering in contemporary contexts. Methods: We used robust ensemble machine learning to train a model using gene microarray data to perform multi-platform predictions including RNA-seq and potentially scRNA-seq. Results: The engineered feature set was composed of many previously reported genes that are associated with patient prognosis. Interestingly, these well-known genes formed a predictive signature only for female patients, and the application of the predictive signature to male patients produced unexpected results. Conclusions: This work demonstrates how annotated “legacy data” can be used to build robust predictive models capable of multi-target predictions across multiple platforms.

Predictive signatures of immune response to vaccination and implications of the immune setpoint remodeling.

In 2020, I featured two articles in the "mSphere of Influence" commentary series that had profound implications for the field of immunology and helped shape my research perspective. These articles were "Global Analyses of Human Immune Variation Reveal Baseline Predictors of Postvaccination Responses" by Tsang et al. (Cell 157:499-513, 2014, https://doi.org/10.1016/j.cell.2014.03.031) and "A crowdsourced analysis to identify ab initio molecular signatures predictive of susceptibility to viral infection" by Fourati et al. (Nat Commun 9:4418, 2018, https://doi.org/10.1038/s41467-018-06735-8). From these topics, the identification of signatures predictive of immune responses to vaccination has greatly advanced and pivoted our understanding of how the immune state at the time of vaccination predicts (and potentially determines) vaccination outcomes. While most of this work has been done using influenza vaccination as a model, pan-vaccine signatures have been also identified. The key implications are their potential use to predict who will respond to vaccinations and to inform strategies for fine-tuning the immune setpoint to enhance immune responses. In addition, investigations in this area led us to understand that immune perturbations, such as acute infections and vaccinations, can remodel the baseline immune state and alter immune responses to future exposures, expanding this exciting field of research. These processes are likely epigenetically encoded, and some examples have already been identified and are discussed in this minireview. Therefore, further research is essential to gain a deeper understanding of how immune exposures modify the epigenome and transcriptome, influence the immune setpoint in response to vaccination, and define its exposure-specific characteristics.

P0024 Circadian regulator-mediated molecular subtypes depict the features of elderly ulcerative colitis

Abstract Background The prevalence of UC in older individuals (i.e., aged > 60 years) is increasing globally because of an ageing population and a subset of population with elderly-onset (i.e., onset aged ≥ 60 years) UC, which causes high burden on healthcare. Circadian rhythms are 24-h oscillations that control various biological processes in the living system and have been recognized to be associated with the development of autoimmune diseases and aging. However, the implication of circadian rhythm in elderly UC remains largely unknown. Methods We downloaded datasets from the Gene Expression Omnibus (GEO) database. The differential expression of circadian regulators between early adult (15-25 years old) and elderly (≥ 60 years old) UC patients was identified. We constructed and compared random forest (RF) and support vector machine (SVM) models. Feature genes were selected to develop a nomogram model with the superior model. The receiver operator characteristic (ROC) curves were used to validate the performances of predictive models and signature genes. The infiltration of 22 kinds of immune cells in elderly UC was analyzed by ssGSEA, and the relationship between feature circadian genes and the proportion of immune infiltration was studied. We identified circadian subtypes based on significantly differentially expressed circadian regulatory genes. Comprehensive evaluation of the two circadian patterns was performed, including immune characteristics and potential functions. Results The expression levels of circadian regulatory genes were extracted and a total of 44 differentially expressed circadian regulators were identified between early adult and elderly UC patients. Five feature circadian regulatory genes, AGT, EGF, NR3C2, PPY, and PYY were determined (p < 0.05) to establish a nomogram model that can predict the incidence of elderly UC with the RF model. The signature circadian regulators were found to be correlated with activated memory CD4+ T cells, gamma delta T cells, and macrophages. We identified two circadian subtypes based on the five significant circadian regulatory genes. Cluster A had a higher immune infiltration of activated CD4+T cells, CD8+ T cells and B cells than cluster B (p < 0.01). The signature circadian genes can also be used to predict the risk of elderly UC and have good diagnostic efficacy. Conclusion The circadian regulatory genes play non-negligible roles in elderly UC. A nomogram model developed by five feature circadian regulatory genes could be used to predict the incidence of elderly UC. Two circadian patterns were identified and evaluated comprehensively, which may provide insights into the classification of elderly UC patients and guide treatment.

A novel ubiquitination-related gene signature for overall survival prediction in patients with liver hepatocellular carcinoma

Liver hepatocellular carcinoma (LIHC) is a highly heterogeneous disease, necessitating the discovery of novel biomarkers to enhance individualized treatment approaches. Recent research has shown the significant involvement of ubiquitin-related genes (UbRGs) in the progression of LIHC. However, the prognostic value of UbRGs in LIHC has not been investigated. In this study, the mRNA expression profiles and clinical data were obtained from public databases of LIHC patients. The least absolute shrinkage and selection operator Cox regression model was employed to construct a multigene signature in the TCGA cohort. Our results showed that a twelve UbRGs signature was developed to categorize patients into two risk groups, with significant differences in expression between LIHC and normal tissues. Patients in the high-risk group exhibited significantly reduced overall survival (OS) and progression-free survival compared to those in the low-risk group. The risk score was identified as an independent predictor for OS in multivariate Cox regression analyses. Receiver operating characteristic curve analysis confirmed the predictive capacity of the signature. Functional analysis revealed enrichment of immune-related pathways and differences in immune status between the two risk groups. The risk score was correlated with 35 transcription factors and 26 eRNA enhancers, and positively associated with tumor mutation burden. Patients in the high-risk group demonstrated decreased sensitivity to targeted and chemotherapeutic drugs than those in the low-risk group. In conclusion, our study identified a twelve UbRGs signature that may serve as a prognostic predictor for LIHC patients and and provide valuable insights for cancer treatment.

Comparative analysis of deep learning and radiomic signatures for overall survival prediction in recurrent high-grade glioma treated with immunotherapy

BackgroundRadiomic analysis of quantitative features extracted from segmented medical images can be used for predictive modeling of prognosis in brain tumor patients. Manual segmentation of the tumor components is time-consuming and poses significant reproducibility issues. We compare the prediction of overall survival (OS) in recurrent high-grade glioma(HGG) patients undergoing immunotherapy, using deep learning (DL) classification networks along with radiomic signatures derived from manual and convolutional neural networks (CNN) automated segmentation.Materials and methodsWe retrospectively retrieved 154 cases of recurrent HGG from multiple centers. Tumor segmentation was performed by expert radiologists and a convolutional neural network (CNN). From the segmented tumors, 2553 radiomic features were extracted for each case. A robust feature subset was selected using intraclass correlation coefficient analysis between manual and automated segmentations. The data was divided into a 9:1 ratio and validated through ten-fold cross-validation and tested on a rotating test set. Features selection was done by the Kruskal–Wallis test. The Radiomics-based OS predictions, generated using Support Vector Machine (SVM), were compared between the two segmentation approaches and against OS prediction by the CNN model adapted for classification. Model efficacy was evaluated using the area under the receiver operating characteristic curve (AUC).ResultsThe clinical model AUC for OS prediction was 0.640 ± 0.013 (mean ± 95% confidence interval) in the training set and 0.610 ± 0.131 in the test set. The radiomics prediction of OS based on manual segmentation outperformed automatic segmentation (AUC of 0.662 ± 0.122 vs. 0.471 ± 0.086, respectively) in the test set. Robust features improved the performance of manual segmentation to AUC of 0.700 ± 0.102, of automated segmentation to 0.554 ± 0.085. The CNN prognosis model demonstrated promising results, with an average AUC of 0.755 ± 0.071 for training sets and 0.700 ± 0.101 for the test set.ConclusionManual segmentation-derived radiomic features outperformed automated segmentation-derived features for predicting OS in recurrent high-grade glioma patients undergoing immunotherapy. The end-to-end CNN prognosis model performed similarly to radiomics modeling using manual-segmentation-derived features without the need for segmentation. The potential time-saving must be weighed against the lower interpretability of end-to-end black box modeling.

Histone H4 lysine 20 monomethylation is not a mark of transcriptional silencers.

Transcriptional silencers are cis -regulatory elements that downregulate the expression of target genes. Although thousands of silencers have been identified experimentally, a predictive chromatin signature of silencers has not been found. H4K20me1 previously was reported to be highly enriched among human silencers, but our reanalysis of those data using an appropriate background revealed that the enrichment is only marginal. We generated H4K20me1 ChIP-seq profiles in Drosophila S2 cells, which similarly showed that H4K20me1 does not mark Drosophila silencers and instead is associated with active transcription. Silencers remain a poorly annotated, difficult to predict class of cis -regulatory elements whose specific chromatin features remain to be identified.

Comparative evaluation of PD-L1 expression and tumor immune microenvironment in molecular subtypes of muscle-invasive bladder cancer and its correlation with survival outcomes.

Immune checkpoint inhibitors have revolutionized treatment of platinum-refractory advanced bladder cancer, offering hope where options are limited. Response varies, however, influenced by factors such as the tumor's immune microenvironment and prior therapy. Muscle-invasive bladder cancer (MIBC) is stratified into molecular subtypes, with distinct clinicopathologic features affecting prognosis and treatment. This study assessed the expression of programmed cell death 1 ligand 1 (PD-L1) and other immune markers in MIBC, categorized by molecular phenotype. Using GATA3 and CK5/6 immunohistochemistry, 90 neoadjuvant chemotherapy-naive MIBC cases were classified into luminal andnon-luminalsubtypes. The immune microenvironment was characterized through immunostaining for PD-L1, CD4, and CD8. We applied PD-L1 positivity thresholds of 1% or greater for tumor cells and 5% or greater for immune cells. Tumors were examined for PD-L1 expression, histologic subtypes, and immune cell infiltration. Varied expression of PD-L1 and T-cell subtype densities were observed among MIBC subtypes. The double-negative subtype displayed the highest PD-L1 immune cell expression and stromal CD4 and CD8 T-cell densities, indicating an active immune profile. The basal subtype exhibited the highest PD-L1 positivity in tumor cells. In contrast, the luminal type showed the lowest PD-L1 tumor and immune cell expression, with high intratumoral CD4 T-cell density. Although PD-L1 expression in tumor or immune cells did not independently affect survival, patients with basal and double-negative tumors had poorer overall survival. This study highlighted the immune diversity of MIBC in the context of molecular subtypes. Distinct molecular and immune profiles could guide the development of predictive signatures for enhanced immunotherapy response in advanced bladder cancer.

Identification of a Novel Immune-Gene Signature with Prognostic Value in Patients with Head and Neck Cancer: A Pilot Study.

The tumor microenvironment has a significant input on prognosis and also for predicting clinical outcomes in various types of cancers. However, tumor tissue is not always available, thus, rendering peripheral blood a preferable alternative in the search for prognostic and predictive gene signatures. Head and neck squamous cell carcinoma (HNSCC) constitutes a quite heterogeneous disease characterized by poor prognosis. Therefore, the discovery of novel therapeutics based on prognostic gene signatures for effective disease governance is of paramount importance. In this study, we report for the first time an immune-gene signature identified in the peripheral blood of HNSCC patients comprising five genes (CLEC4C, IL23A, LCK, LY9, and CD19) which were more than threefold downregulated as compared to healthy individuals and were associated with poor prognosis. By performing analyses of HNSCC tumor samples from The Cancer Genome Atlas (TCGA) database, we discovered that decreased expression of these genes, both as single genes and as a 5-gene signature (5-GS), was significantly correlated with worse overall survival (OS). Our data show that the levels of expression of the 5-GS represent an immune profile predicting OS in patients with HNSCC.

A Novel Disulfidptosis-Related Risk Signature for Prognostic Prediction in Patients With Ewing Sarcoma.

Ewing sarcoma (ES) is a malignant bone tumor prevalent among children and adolescents. Disulfidptosis represents a novel form of cell death; however, the mechanism of disulfidptosis in ES remains unclear. Our aim is to explore the disulfidptosis-related prognostic signature in ES. Utilizing transcriptomic and clinical data of ES, disulfidptosis-related hub genes (DRHGs) were identified by differential gene expression analysis and Least Absolute Shrinkage and Selection Operator (LASSO) Cox regression analysis. A disulfidptosis-related risk score model (DRRS) was constructed based on these DRHGs. The performance of DRRS was assessed using survival analysis and receiver operating characteristic curve analysis. Immune cell infiltration in different risk subgroups and correlations between DRRS and antitumor reagents were also analyzed. In this study, we developed a disulfidptosis-related prognostic feature based on LRPPRC (leucine rich pentatricopeptide repeat containing), IQGAP1 (IQ motif containing GTPase activating protein 1), NDUFS1 (NADH:ubiquinone oxidoreductase core subunit S1), and TLN1 (talin 1), which may serve as a predictive and independent risk factor for ES. ES patients in the high-risk group exhibited a poorer prognosis, had a higher proportion of myeloid-derived suppressor cells (MDSCs) and M2 type of tumor-associated macrophages, and showed heightened sensitivity to some antitumor agents such as nilotinib and olaparib. This study is the first to construct a disulfidptosis-related prognostic signature that may predict the prognosis and immune response in ES patients, thereby providing a new reference for understanding the mechanisms of ES and guiding immunotherapy.

Learning predictive signatures of HLA type from T-cell repertoires.

T cells recognize a wide range of pathogens using surface receptors that interact directly with peptides presented on major histocompatibility complexes (MHC) encoded by the HLA loci in humans. Understanding the association between T cell receptors (TCR) and HLA alleles is an important step towards predicting TCR-antigen specificity from sequences. Here we analyze the TCR alpha and beta repertoires of large cohorts of HLA-typed donors to systematically infer such associations, by looking for overrepresentation of TCRs in individuals with a common allele.TCRs, associated with a specific HLA allele, exhibit sequence similarities that suggest prior antigen exposure. Immune repertoire sequencing has produced large numbers of datasets, however the HLA type of the corresponding donors is rarely available. Using our TCR-HLA associations, we trained a computational model to predict the HLA type of individuals from their TCR repertoire alone. We propose an iterative procedure to refine this model by using data from large cohorts of untyped individuals, by recursively typing them using the model itself. The resulting model shows good predictive performance, even for relatively rare HLA alleles.

Deciphering functional tumor-immune crosstalk through highly multiplexed imaging and deep visual proteomics.

Deciphering the intricate tumor-immune interactions within the microenvironment is crucial for advancing cancer immunotherapy. Here, we introduce mipDVP, an advanced approach integrating highly multiplexed imaging, single-cell laser microdissection, and sensitive mass spectrometry to spatially profile the proteomes of distinct cell populations in a human colorectal and tonsil cancer with high sensitivity. In a colorectal tumor-a representative cold tumor-we uncovered spatial compartmentalization of an immunosuppressive macrophage barrier that potentially impedes Tcell infiltration. Spatial proteomic analysis revealed distinct functional states of Tcells in different tumor compartments. In a tonsil cancer sample-a hot tumor-we identified significant proteomic heterogeneity among cells influenced by proximity to cytotoxic Tcell subtypes. Tcells in the tumor parenchyma exhibit metabolic adaptations to hypoxic regions. Our spatially resolved, highly multiplexed strategy deciphers the complex cellular interplay within the tumor microenvironment, offering valuable insights for identifying immunotherapy targets and predictive signatures.

Analysis and Validation of Tyrosine Metabolism-related Prognostic Features for Liver Hepatocellular Carcinoma Therapy.

To explore tyrosine metabolism-related characteristics in liver hepatocellular carcinoma (LIHC) and to establish a risk signature for the prognostic prediction of LIHC. Novel prognostic signatures contribute to the mining of novel biomarkers, which are essential for the construction of a precision medicine system for LIHC and the improvement of survival. Tyrosine metabolism plays a critical role in the initiation and development of LIHC. Based on the tyrosine metabolism-related characteristics in LIHC, this study developed a risk signature to improve the prognostic prediction of patients with LIHC. To investigate the correlation between tyrosine metabolism and progression of LIHC and to develop a tyrosine metabolism-related prognostic model. Gene expression and clinicopathological information of LIHC were obtained from The Cancer Genome Atlas (TCGA) database. Distinct subtypes of LIHC were classified by performing consensus cluster analysis on the tyrosine metabolism-related genes. Univariate and Lasso Cox regression were used to develop a RiskScore prognosis model. Kaplan-Meier (KM) survival analysis with log-rank test and area under the curve (AUC) of receiver operating characteristic (ROC) were employed in the prognostic evaluation and prediction validation. Immune infiltration, tyrosine metabolism score, and pathway enrichment were evaluated using single-sample gene set enrichment analysis (ssGSEA). Finally, a nomogram model was developed with the RiskScore and other clinicopathological features. Based on the tyrosine metabolism genes in the TCGA cohort, we identified 3 tyrosine metabolism-related subtypes showing significant prognostic differences. Four candidate genes selected from the common differentially expressed genes (DEGs) between the 3 subtypes were used to develop a RiskScore model, which could effectively divide LIHC patients into high- and lowrisk groups. In both the training and validation sets, high-risk patients tended to have worse overall survival, less active immunotherapy response, higher immune infiltration and clinical grade, and higher oxidative, fatty, and xenobiotic metabolism pathways. Multivariate analysis confirmed that the RiskScore was an independent indicator for the prognosis of LIHC. The results from pan-- cancer analysis also supported that the RiskScore had a strong prognostic performance in other cancers. The nomogram demonstrated that the RiskScore contributed the most to the prediction of LIHC prognosis. Our study developed a tyrosine metabolism-related risk model that performed well in survival prediction, showing the potential to serve as an independent prognostic predictor for LIHC treatment.

A novel combined oxidative stress and extracellular matrix related predictive gene signature for keratoconus.

Keratoconus (KC) is an ectatic cornea disease with high prevalence and asymptomatic at early stage, leading to decreased visual acuity and even cornea transplantation. However, the etiology mechanism of keratoconus is still poorly understood. Oxidative stress (OS) and extracellular matrix (ECM) remodeling play critical roles in keratoconus development. Here, based on keratoconus datasets from GEO database, we obtained 454 differentially expressed genes (DEGs), which were further intersected with oxidative stress (OS) and extracellular matrix (ECM) genesets from MSigDB database. A total of 17 OS- and ECM-related DEGs (OEDEGs) were identified. Feature genes were screened by least absolute shrinkage and selection operator (LASSO) and support vector machine recursive feature elimination (SVM-RFE) algorithms, and a six-gene (COL1A1, CYP1B1, MMP3, HMOX1, FOS and GDF15) classification model was developed utilizing Logistic regression (LR), Support Vector Machine (SVM) and Naïve Bayes (NB) algorithms respectively, which was further verified in internal and external cohort. Subsequently, a predictive nomogram was constructed for KC patients. Six signature genes showed a strong correlation with the infiltration level of macrophages M1, neutrophils and eosinophils. Additionally, in vitro qRT-PCR validated the decreased expression of signature genes in either keratoconus clinical samples or human cornea epithelial (HCE) cells grown on soft hydrogel substrate. Finally, we revealed that CYP1B1 and GDF15 regulate cellular proliferation and response to oxidative stress. In conclusion, the developed combined OS and ECM gene signature showed excellent performance for keratoconus prediction, providing beneficial perspectives for keratoconus pathogenesis.