A novel inflammation‑associated prognostic signature for clear cell renal cell carcinoma

Comprehensive characterization of stemness-related lncRNAs in triple-negative breast cancer identified a novel prognostic signature related to treatment outcomes, immune landscape analysis and therapeutic guidance: a silico analysis with in vivo experiments

BACKGROUND

Cancer stem cells (CSCs) and long non-coding RNAs (lncRNAs) are known to play a crucial role in the growth, migration, recurrence, and drug resistance of tumor cells, particularly in triple-negative breast cancer (TNBC). This study aims to investigate stemness-related lncRNAs (SRlncRNAs) as potential prognostic indicators for TNBC patients.

METHODS

Utilizing RNA sequencing data and corresponding clinical information from the TCGA database, and employing Weighted Gene Co-expression Network Analysis (WGCNA) on TNBC mRNAsi sourced from an online database, stemness-related genes (SRGs) and SRlncRNAs were identified. A prognostic model was developed using univariate Cox and LASSO-Cox analysis based on SRlncRNAs. The performance of the model was evaluated using Kaplan-Meier analysis, ROC curves, and ROC-AUC. Additionally, the study delved into the underlying signaling pathways and immune status associated with the divergent prognoses of TNBC patients.

RESULTS

The research identified a signature of six SRlncRNAs (AC245100.6, LINC02511, AC092431.1, FRGCA, EMSLR, and MIR193BHG) for TNBC. Risk scores derived from this signature were found to correlate with the abundance of plasma cells. Furthermore, the nominated chemotherapy drugs for TNBC exhibited considerable variability between different risk score groups. RT-qPCR validation confirmed abnormal expression patterns of these SRlncRNAs in TNBC stem cells, affirming the potential of the SRlncRNAs signature as a prognostic biomarker.

CONCLUSION

The identified signature not only demonstrates predictive power in terms of patient outcomes but also provides insights into the underlying biology, signaling pathways, and immune status associated with TNBC prognosis. The findings suggest the possibility of guiding personalized treatments, including immune checkpoint gene therapy and chemotherapy strategies, based on the risk scores derived from the SRlncRNA signature. Overall, this research contributes valuable knowledge towards advancing precision medicine in the context of TNBC.

Decoding the tumor microenvironment and molecular mechanism: unraveling cervical cancer subpopulations and prognostic signatures through scRNA-Seq and bulk RNA-seq analyses

Background

Cervical carcinoma (CC) represents a prevalent gynecological neoplasm, with a discernible rise in prevalence among younger cohorts observed in recent years. Nonetheless, the intrinsic cellular heterogeneity of CC remains inadequately investigated.

Methods

We utilized single-cell RNA sequencing (scRNA-seq) transcriptomic analysis to scrutinize the tumor epithelial cells derived from four specimens of cervical carcinoma (CC) patients. This method enabled the identification of pivotal subpopulations of tumor epithelial cells and elucidation of their contributions to CC progression. Subsequently, we assessed the influence of associated molecules in bulk RNA sequencing (Bulk RNA-seq) cohorts and performed cellular experiments for validation purposes.

Results

Through our analysis, we have discerned C3 PLP2+ Tumor Epithelial Progenitor Cells as a noteworthy subpopulation in cervical carcinoma (CC), exerting a pivotal influence on the differentiation and progression of CC. We have established an independent prognostic indicator-the PLP2+ Tumor EPCs score. By stratifying patients into high and low score groups based on the median score, we have observed that the high-score group exhibits diminished survival rates compared to the low-score group. The correlations observed between these groups and immune infiltration, enriched pathways, single-nucleotide polymorphisms (SNPs), drug sensitivity, among other factors, further underscore their impact on CC prognosis. Cellular experiments have validated the significant impact of ATF6 on the proliferation and migration of CC cell lines.

Conclusion

This study enriches our comprehension of the determinants shaping the progression of CC, elevates cognizance of the tumor microenvironment in CC, and offers valuable insights for prospective CC therapies. These discoveries contribute to the refinement of CC diagnostics and the formulation of optimal therapeutic approaches.

Molecular mechanisms of pancreatic cancer liver metastasis: the role of PAK2

Background

Pancreatic cancer remains an extremely malignant digestive tract tumor, posing a significant global public health burden. Patients with pancreatic cancer, once metastasis occurs, lose all hope of cure, and prognosis is extremely poor. It is important to investigate liver metastasis of Pancreatic cancer in depth, not just because it is the most common form of metastasis in pancreatic cancer, but also because it is crucial for treatment planning and prognosis assessment. This study aims to delve into the mechanisms of pancreatic cancer liver metastasis, with the goal of providing crucial scientific groundwork for the development of future treatment methods and drugs.

Methods

We explored the mechanisms of pancreatic cancer liver metastasis using single-cell sequencing data (GSE155698 and GSE154778) and bulk data (GSE71729, GSE19279, TCGA-PAAD). Initially, Seurat package was employed for single-cell data processing to obtain expression matrices for primary pancreatic cancer lesions and liver metastatic lesions. Subsequently, high-dimensional weighted gene co-expression network analysis (hdWGCNA) was used to identify genes associated with liver metastasis. Machine learning algorithms and COX regression models were employed to further screen genes related to patient prognosis. Informed by both biological understanding and the outcomes of algorithms, we meticulously identified the ultimate set of liver metastasis-related gene (LRG). In the study of LRG genes, various databases were utilized to validate their association with pancreatic cancer liver metastasis. In order to analyze the effects of these agents on tumor microenvironment, we conducted an in-depth analysis, including changes in signaling pathways (GSVA), cell differentiation (pseudo-temporal analysis), cell communication networks (cell communication analysis), and downstream transcription factors (transcription factor activity prediction). Additionally, drug sensitivity analysis and metabolic analysis were performed to reveal the effects of LRG on gemcitabine resistance and metabolic pathways. Finally, functional experiments were conducted by silencing the expression of LRG in PANC-1 and Bx-PC-3 cells to validate its influence to proliferation and invasiveness on PANC-1 and Bx-PC-3 cells.

Results

Through a series of algorithmic filters, we identified PAK2 as a key gene promoting pancreatic cancer liver metastasis. GSVA analysis elucidated the activation of the TGF-beta signaling pathway by PAK2 to promote the occurrence of liver metastasis. Pseudo-temporal analysis revealed a significant correlation between PAK2 expression and the lower differentiation status of pancreatic cancer cells. Cell communication analysis revealed that overexpression of PAK2 promotes communication between cancer cells and the tumor microenvironment. Transcription factor activity prediction displayed the transcription factor network regulated by PAK2. Drug sensitivity analysis and metabolic analysis revealed the impact of PAK2 on gemcitabine resistance and metabolic pathways. CCK8 experiments showed that silencing PAK2 led to a decrease in the proliferative capacity of pancreatic cancer cells and scratch experiments demonstrated that low expression of PAK2 decreased invasion capability in pancreatic cancer cells. Flow cytometry reveals that PAK2 significantly inhibited apoptosis in pancreatic cancer cell lines. Molecules related to the TGF-beta pathway decreased with the inhibition of PAK2, and there were corresponding significant changes in molecules associated with EMT.

Conclusion

PAK2 facilitated the angiogenic potential of cancer cells and promotes the epithelial-mesenchymal transition process by activating the TGF-beta signaling pathway. Simultaneously, it decreased the differentiation level of cancer cells, consequently enhancing their malignancy. Additionally, PAK2 fostered communication between cancer cells and the tumor microenvironment, augments cancer cell chemoresistance, and modulates energy metabolism pathways. In summary, PAK2 emerged as a pivotal gene orchestrating pancreatic cancer liver metastasis. Intervening in the expression of PAK2 may offer a promising therapeutic strategy for preventing liver metastasis of pancreatic cancer and improving its prognosis.

Machine learning-enhanced insights into sphingolipid-based prognostication: revealing the immunological landscape and predictive proficiency for immunomotherapy and chemotherapy responses in pancreatic carcinoma

Background: With a poor prognosis for affected individuals, pancreatic adenocarcinoma (PAAD) is known as a complicated and diverse illness. Immunocytes have become essential elements in the development of PAAD. Notably, sphingolipid metabolism has a dual function in the development of tumors and the invasion of the immune system. Despite these implications, research on the predictive ability of sphingolipid variables for PAAD prognosis is strikingly lacking, and it is yet unclear how they can affect PAAD immunotherapy and targeted pharmacotherapy. Methods: The investigation process included SPG detection while also being pertinent to the prognosis for PAAD. Both the analytical capability of CIBERSORT and the prognostic capability of the pRRophetic R package were used to evaluate the immunological environments of the various HCC subtypes. In addition, CCK-8 experiments on PAAD cell lines were carried out to confirm the accuracy of drug sensitivity estimates. The results of these trials, which also evaluated cell survival and migratory patterns, confirmed the usefulness of sphingolipid-associated genes (SPGs). Results: As a result of this thorough investigation, 32 SPGs were identified, each of which had a measurable influence on the dynamics of overall survival. This collection of genes served as the conceptual framework for the development of a prognostic model, which was carefully assembled from 10 chosen genes. It should be noted that this grouping of patients into cohorts with high and low risk was a sign of different immune profiles and therapy responses. The increased abundance of SPGs was identified as a possible sign of inadequate responses to immune-based treatment approaches. The careful CCK-8 testing carried out on PAAD cell lines was of the highest importance for providing clear confirmation of drug sensitivity estimates. Conclusion: The significance of Sphingolipid metabolism in the complex web of PAAD development is brought home by this study. The novel risk model, built on the complexity of sphingolipid-associated genes, advances our understanding of PAAD and offers doctors a powerful tool for developing personalised treatment plans that are specifically suited to the unique characteristics of each patient.

Unveiling the role of regulatory T cells in the tumor microenvironment of pancreatic cancer through single-cell transcriptomics and in vitro experiments

Background

In order to investigate the impact of Treg cell infiltration on the immune response against pancreatic cancer within the tumor microenvironment (TME), and identify crucial mRNA markers associated with Treg cells in pancreatic cancer, our study aims to delve into the role of Treg cells in the anti-tumor immune response of pancreatic cancer.

Methods

The ordinary transcriptome data for this study was sourced from the GEO and TCGA databases. It was analyzed using single-cell sequencing analysis and machine learning. To assess the infiltration level of Treg cells in pancreatic cancer tissues, we employed the CIBERSORT method. The identification of genes most closely associated with Treg cells was accomplished through the implementation of weighted gene co-expression network analysis (WGCNA). Our analysis of single-cell sequencing data involved various quality control methods, followed by annotation and advanced analyses such as cell trajectory analysis and cell communication analysis to elucidate the role of Treg cells within the pancreatic cancer microenvironment. Additionally, we categorized the Treg cells into two subsets: Treg1 associated with favorable prognosis, and Treg2 associated with poor prognosis, based on the enrichment scores of the key genes. Employing the hdWGCNA method, we analyzed these two subsets to identify the critical signaling pathways governing their mutual transformation. Finally, we conducted PCR and immunofluorescence staining in vitro to validate the identified key genes.

Results

Based on the results of immune infiltration analysis, we observed significant infiltration of Treg cells in the pancreatic cancer microenvironment. Subsequently, utilizing the WGCNA and machine learning algorithms, we ultimately identified four Treg cell-related genes (TRGs), among which four genes exhibited significant correlations with the occurrence and progression of pancreatic cancer. Among them, CASP4, TOB1, and CLEC2B were associated with poorer prognosis in pancreatic cancer patients, while FYN showed a correlation with better prognosis. Notably, significant differences were found in the HIF-1 signaling pathway between Treg1 and Treg2 cells identified by the four genes. These conclusions were further validated through in vitro experiments.

Conclusion

Treg cells played a crucial role in the pancreatic cancer microenvironment, and their presence held a dual significance. Recognizing this characteristic was vital for understanding the limitations of Treg cell-targeted therapies. CASP4, FYN, TOB1, and CLEC2B exhibited close associations with infiltrating Treg cells in pancreatic cancer, suggesting their involvement in Treg cell functions. Further investigation was warranted to uncover the mechanisms underlying these associations. Notably, the HIF-1 signaling pathway emerged as a significant pathway contributing to the duality of Treg cells. Targeting this pathway could potentially revolutionize the existing treatment approaches for pancreatic cancer.

Revealing the role of regulatory T cells in the tumor microenvironment of lung adenocarcinoma: a novel prognostic and immunotherapeutic signature

Background

Regulatory T cells (Tregs), are a key class of cell types in the immune system. In the tumor microenvironment (TME), the presence of Tregs has important implications for immune response and tumor development. Relatively little is known about the role of Tregs in lung adenocarcinoma (LUAD).

Methods

Tregs were identified using but single-cell RNA sequencing (scRNA-seq) analysis and interactions between Tregs and other cells in the TME were investigated. Next, we used multiple bulk RNA-seq datasets to construct risk models based on marker genes of Tregs and explored differences in prognosis, mutational landscape, immune cell infiltration and immunotherapy between high- and low-risk groups, and finally, qRT-PCR and cell function experiments were performed to validate the model genes.

Results

The cellchat analysis showed that MIF-(CD74+CXCR4) pairs play a key role in the interaction of Tregs with other cell subpopulations, and the Tregs-associated signatures (TRAS) could well classify multiple LUAD cohorts into high- and low-risk groups. Immunotherapy may offer greater potential benefits to the low-risk group, as indicated by their superior survival, increased infiltration of immune cells, and heightened expression of immune checkpoints. Finally, the experiment verified that the model genes LTB and PTTG1 were relatively highly expressed in cancer tissues, while PTPRC was relatively highly expressed in paracancerous tissues. Colony Formation assay confirmed that knockdown of PTTG1 reduced the proliferation ability of LUAD cells.

Conclusion

TRAS were constructed using scRNA-seq and bulk RNA-seq to distinguish patient risk subgroups, which may provide assistance in the clinical management of LUAD patients.

GADD45B regulates the carcinogenesis process of chronic atrophic gastritis and the metabolic pathways of gastric cancer

Background

Gastric cancer continues to be a significant global healthcare challenge, and its burden remains substantial. The development of gastric cancer (GC) is closely linked to chronic atrophic gastritis (CAG), yet there is a scarcity of research exploring the underlying mechanisms of CAG-induced carcinogenesis.

Methods

In this study, we conducted a comprehensive investigation into the oncogenes involved in CAG using both bulk transcriptome and single-cell transcriptome data. Our approach employed hdWGCNA to identify pathogenic genes specific to CAG, with non-atrophic gastritis (NAG) serving as the control group. Additionally, we compared CAG with GC, using normal gastric tissue as the control group in the single-cell transcriptome analysis. By intersecting the identified pathogenic genes, we pinpointed key network molecules through protein interaction network analysis. To further refine the gene selection, we applied LASSO, SVM-RFE, and RF techniques, which resulted in a set of cancer-related genes (CRGs) associated with CAG. To identify CRGs potentially linked to gastric cancer progression, we performed a univariate COX regression analysis on the gene set. Subsequently, we explored the relationship between CRGs and immune infiltration, drug sensitivity, and clinical characteristics in gastric cancer patients. We employed GSVA to investigate how CRGs regulated signaling pathways in gastric cancer cells, while an analysis of cell communication shed light on the impact of CRGs on signal transmission within the gastric cancer tumor microenvironment. Lastly, we analyzed changes in metabolic pathways throughout the progression of gastric cancer.

Results

Using hdWGCNA, we have identified a total of 143 pathogenic genes that were shared by CAG and GC. To further investigate the underlying mechanisms, we conducted protein interaction network analysis and employed machine learning screening techniques. As a result, we have identified 15 oncogenes that are specifically associated with chronic atrophic gastritis. By performing ROC reanalysis and prognostic analysis, we have determined that GADD45B is the most significant gene involved in the carcinogenesis of CAG. Immunohistochemical staining and differential analysis have revealed that GADD45B expression was low in GC tissues while high in normal gastric tissues. Moreover, based on prognostic analysis, high expression of GADD45B has been correlated with poor prognosis in GC patients. Additionally, an analysis of immune infiltration has shown a relationship between GADD45B and the infiltration of various immune cells. By correlating GADD45B with clinical characteristics, we have found that it primarily affects the depth of invasion in GC. Through cell communication analysis, we have discovered that the CD99 signaling pathway network and the CDH signaling pathway network are the main communication pathways that significantly alter the microenvironment of gastric tissue during the development of chronic atrophic gastritis. Specifically, GADD45B-low GC cells were predominantly involved in the network communication of the CDH signaling pathway, while GADD45B-high GC cells played a crucial role in both signaling pathways. Furthermore, we have identified several metabolic pathways, including D-Glutamine and D-glutamate metabolism and N-Glycan biosynthesis, among others, that played important roles in the occurrence and progression of GC, in addition to the six other metabolic pathways. In summary, our study highlighted the discovery of 143 pathogenic genes shared by CAG and GC, with a specific focus on 15 oncogenes associated with CAG. We have identified GADD45B as the most important gene in the carcinogenesis of CAG, which exhibited differential expression in GC tissues compared to normal gastric tissues. Moreover, GADD45B expression was correlated with patient prognosis and is associated with immune cell infiltration. Our findings also emphasized the impact of the CD99 and CDH signaling pathway networks on the microenvironment of gastric tissue during the development of CAG. Additionally, we have identified key metabolic pathways involved in GC progression.

Conclusion

GADD45B, an oncogene implicated in chronic atrophic gastritis, played a critical role in GC development. Decreased expression of GADD45B was associated with the onset of GC. Moreover, GADD45B expression levels were closely tied to poor prognosis in GC patients, influencing the infiltration patterns of various cells within the tumor microenvironment, as well as impacting the metabolic pathways involved in GC progression.

Unveiling efferocytosis-related signatures through the integration of single-cell analysis and machine learning: a predictive framework for prognosis and immunotherapy response in hepatocellular carcinoma

Background

Hepatocellular carcinoma (HCC) represents a prominent gastrointestinal malignancy with a grim clinical outlook. In this regard, the discovery of novel early biomarkers holds substantial promise for ameliorating HCC-associated mortality. Efferocytosis, a vital immunological process, assumes a central position in the elimination of apoptotic cells. However, comprehensive investigations exploring the role of efferocytosis-related genes (EFRGs) in HCC are sparse, and their regulatory influence on HCC immunotherapy and targeted drug interventions remain poorly understood.

Methods

RNA sequencing data and clinical characteristics of HCC patients were acquired from the TCGA database. To identify prognostically significant genes in HCC, we performed the limma package and conducted univariate Cox regression analysis. Subsequently, machine learning algorithms were employed to identify hub genes. To assess the immunological landscape of different HCC subtypes, we employed the CIBERSORT algorithm. Furthermore, single-cell RNA sequencing (scRNA-seq) was utilized to investigate the expression levels of ERFGs in immune cells and to explore intercellular communication within HCC tissues. The migratory capacity of HCC cells was evaluated using CCK-8 assays, while drug sensitivity prediction reliability was determined through wound-healing assays.

Results

We have successfully identified a set of nine genes, termed EFRGs, that hold significant potential for the establishment of a hepatocellular carcinoma-specific prognostic model. Furthermore, leveraging the individual risk scores derived from this model, we were able to stratify patients into two distinct risk groups, unveiling notable disparities in terms of immune infiltration patterns and response to immunotherapy. Notably, the model's capacity to accurately predict drug responses was substantiated through comprehensive experimental investigations, encompassing wound-healing assay, and CCK8 experiments conducted on the HepG2 and Huh7 cell lines.

Conclusions

We constructed an EFRGs model that serves as valuable tools for prognostic assessment and decision-making support in the context of immunotherapy and chemotherapy.

Efferocytosis signatures as prognostic markers for revealing immune landscape and predicting immunotherapy response in hepatocellular carcinoma

Background: Hepatocellular carcinoma (HCC) is a highly lethal liver cancer with late diagnosis; therefore, the identification of new early biomarkers could help reduce mortality. Efferocytosis, a process in which one cell engulfs another cell, including macrophages, dendritic cells, NK cells, etc., plays a complex role in tumorigenesis, sometimes promoting and sometimes inhibiting tumor development. However, the role of efferocytosis-related genes (ERGs) in HCC progression has been poorly studied, and their regulatory effects in HCC immunotherapy and drug targeting have not been reported. Methods: We downloaded efferocytosis-related genes from the Genecards database and screened for ERGs that showed significant expression changes between HCC and normal tissues and were associated with HCC prognosis. Machine learning algorithms were used to study prognostic gene features. CIBERSORT and pRRophetic R packages were used to evaluate the immune environment of HCC subtypes and predict treatment response. CCK-8 experiments conducted on HCC cells were used to assess the reliability of drug sensitivity prediction. Results: We constructed a prognostic prediction model composed of six genes, and the ROC curve showed good predictive accuracy of the risk model. In addition, two ERG-related subgroups in HCC showed significant differences in tumor immune landscape, immune response, and prognostic stratification. The CCK-8 experiment conducted on HCC cells confirmed the reliability of drug sensitivity prediction. Conclusion: Our study emphasizes the importance of efferocytosis in HCC progression. The risk model based on efferocytosis-related genes developed in our study provides a novel precision medicine approach for HCC patients, allowing clinicians to customize treatment plans based on unique patient characteristics. The results of our investigation carry noteworthy implications for the development of individualized treatment approaches involving immunotherapy and chemotherapy, thereby potentially facilitating the realization of personalized and more efficacious therapeutic interventions for HCC.

Multi-omic analysis of glycolytic signatures: exploring the predictive significance of heterogeneity and stemness in immunotherapy response and outcomes in hepatocellular carcinoma

Background: Hepatocellular carcinoma (HCC) is a global health challenge with complex pathophysiology, characterized by high mortality rates and poor early detection due to significant tumor heterogeneity. Stemness significantly contributes to the heterogeneity of HCC tumors, and glycolysis is crucial for maintaining stemness. However, the predictive significance of glycolysis-related metabolic genes (GMGs) in HCC remains unknown. Therefore, this study aimed to identify critical GMGs and establish a reliable model for HCC prognosis. Methods: GMGs associated with prognosis were identified by evaluating genes with notable expression changes between HCC and normal tissues retrieved from the MsigDB database. Prognostic gene characteristics were established using univariate and multivariate Cox regression studies for prognosis prediction and risk stratification. The "CIBERSORT" and "pRRophetic" R packages were respectively used to evaluate the immunological environment and predict treatment response in HCC subtypes. The HCC stemness score was obtained using the OCLR technique. The precision of drug sensitivity prediction was evaluated using CCK-8 experiments performed on HCC cells. The miagration and invasion ability of HCC cell lines with different riskscores were assessed using Transwell and wound healing assays. Results: The risk model based on 10 gene characteristics showed high prediction accuracy as indicated by the receiver operating characteristic (ROC) curves. Moreover, the two GMG-related subgroups showed considerable variation in the risk of HCC with respect to tumor stemness, immune landscape, and prognostic stratification. The in vitro validation of the model's ability to predict medication response further demonstrated its reliability. Conclusion: Our study highlights the importance of stemness variability and inter-individual variation in determining the HCC risk landscape. The risk model we developed provides HCC patients with a novel method for precision medicine that enables clinical doctors to customize treatment plans based on unique patient characteristics. Our findings have significant implications for tailored immunotherapy and chemotherapy methods, and may pave the way for more personalized and effective treatment strategies for HCC.

Ensemble deep learning enhanced with self-attention for predicting immunotherapeutic responses to cancers

Introduction

Despite the many benefits immunotherapy has brought to patients with different cancers, its clinical applications and improvements are still hindered by drug resistance. Fostering a reliable approach to identifying sufferers who are sensitive to certain immunotherapeutic agents is of great clinical relevance.

Methods

We propose an ELISE (Ensemble Learning for Immunotherapeutic Response Evaluation) pipeline to generate a robust and highly accurate approach to predicting individual responses to immunotherapies. ELISE employed iterative univariable logistic regression to select genetic features of patients, using Monte Carlo Tree Search (MCTS) to tune hyperparameters. In each trial, ELISE selected multiple models for integration based on add or concatenate stacking strategies, including deep neural network, automatic feature interaction learning via self-attentive neural networks, deep factorization machine, compressed interaction network, and linear neural network, then adopted the best trial to generate a final approach. SHapley Additive exPlanations (SHAP) algorithm was applied to interpret ELISE, which was then validated in an independent test set.

Result

Regarding prediction of responses to atezolizumab within esophageal adenocarcinoma (EAC) patients, ELISE demonstrated a superior accuracy (Area Under Curve [AUC] = 100.00%). AC005786.3 (Mean [|SHAP value|] = 0.0097) was distinguished as the most valuable contributor to ELISE output, followed by SNORD3D (0.0092), RN7SKP72 (0.0081), EREG (0.0069), IGHV4-80 (0.0063), and MIR4526 (0.0063). Mechanistically, immunoglobulin complex, immunoglobulin production, adaptive immune response, antigen binding and others, were downregulated in ELISE-neg EAC subtypes and resulted in unfavorable responses. More encouragingly, ELISE could be extended to accurately estimate the responsiveness of various immunotherapeutic agents against other cancers, including PD1/PD-L1 suppressor against metastatic urothelial cancer (AUC = 88.86%), and MAGE-A3 immunotherapy against metastatic melanoma (AUC = 100.00%).

Discussion

This study presented deep insights into integrating ensemble deep learning with self-attention as a mechanism for predicting immunotherapy responses to human cancers, highlighting ELISE as a potential tool to generate reliable approaches to individualized treatment.