SEARCH · PubMed Health

Results for “cancer subtype classification”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 recordsLinked to original sources

PubMed2026

A nonlinear multi-omics data integration and classification model based on pathway self-attention and graph convolutional networks.

The abundance of omics data has significantly advanced the development of multi-omics data integration techniques. Non-linear embedding approaches for data integration have gradually become the mainstream in multi-omics research, as these approaches can substantially improve cancer analysis by enhancing the quality of the embeddings. However, current multi-omics data integration methods are typically confined to omics measurements, neglecting domain-specific prior knowledge encompassing biological pathways. In this study, we proposed a multi-omics integrated classification model, PathTransGCN, based on pathway self-attention and graph convolutional networks (GCN). The model integrated biological pathway information into multi-omics data analysis with the aim of enhancing the accuracy of cancer classification. Multi-omics data for breast cancer (BRCA), non-small cell lung cancer (NSCLC), and low-grade glioma (LGG) were obtained from The Cancer Genome Atlas (TCGA) and UCSC Xena databases. These data included gene mutations, DNA methylation, copy number variations, and gene expression, and were used to assess the model's generalizability across different cancers. First, PathTransGCN employed a pathway self-attention module to learn latent representations of samples across different pathways, thereby obtaining multi-omics integration vectors. Concurrently, a patient similarity network (PSN) was constructed using the similarity network fusion (SNF) approach. Second, the integrated vectors and the PSN were jointly fed into a GCN for end-to-end training, enabling precise classification of cancer subtypes. Through multi-omics data analysis of the BRCA dataset, PathTransGCN outperformed several popular algorithms (such as MoGCN and DeePathNet) in the five-class classification of cancer subtypes, achieving an accuracy rate of 87.6% and an F1 score of 86.4%. Moreover, the model demonstrated robust generalization capabilities across both NSCLC and LGG datasets, while effectively identifying key disease-associated biomarkers at the pathway level. Experimental results demonstrate that PathTransGCN exhibits outstanding performance in integrating omics data and delivering interpretable classification outcomes, presenting significant potential for clinical applications.

Humans↗

PubMed2025

IGCN: integrative graph convolution networks for patient level insights and biomarker discovery in multi-omics integration.

MOTIVATION: Developing computational tools for integrative analysis across multiple types of omics data has been of immense importance in cancer molecular biology and precision medicine research. While recent advancements have yielded integrative prediction solutions for multi-omics data, these methods lack a comprehensive and cohesive understanding of the rationale behind their specific predictions. To shed light on personalized medicine and unravel previously unknown characteristics within integrative analysis of multi-omics data, we introduce a novel integrative neural network approach for cancer molecular subtype and biomedical classification applications, named Integrative Graph Convolutional Networks (IGCN). RESULTS: To demonstrate the superiority of IGCN, we compare its performance with other state-of-the-art approaches across different cancer subtype and biomedical classification tasks. Our experimental results show that our proposed model outperforms the state-of-the-art and baseline methods. IGCN identifies which types of omics data receive more emphasis for each patient when predicting a specific class. Additionally, IGCN has the capability to pinpoint significant biomarkers from a range of omics data types. AVAILABILITY AND IMPLEMENTATION: The source code is available at https://github.com/bozdaglab/IGCN.

Humans↗

PubMed1994

Detection of keratin subtypes in routinely processed cervical tissue: implications for tumour classification and the study of cervix cancer aetiology.

We investigated the expression of keratin subtypes 7, 8, 10, 13, 14, 17, 18 and 19 in the normal cervix, in cervical intraepithelial neoplasia (CIN) lesions and in cervical carcinomas, using a selected panel of monoclonal keratin antibodies, reactive with routinely processed, formalin fixed paraffin embedded tissue fragments. The reaction patterns derived for each keratin antibody were compared with known expression patterns of the various epithelia, previously examined in frozen tissues. Although the reactivity of the antibodies was generally acceptable, considerable modifications to the manufacturers' staining instructions were often necessary. For some antibodies, which were previously thought to be reactive with fresh frozen tissue only, we developed staining protocols rendering them reactive with routinely processed material. As with previous findings in frozen sections we observed increasing expression of keratins 7, 8, 17, 18 and 19 with increasing grade of CIN. In cervical carcinomas the differences in keratin detectability between the main categories were more pronounced than in frozen sections, probably due to fixation and processing. For routine pathology, keratin phenotyping of cervical lesions may be of value in classification. The fact that keratin 7 was detected for the first time in reserve cells, and that this keratin was also found to be expressed in a considerable number of CIN lesions and cervical carcinomas supports the suggestion that reserve cells are a common progenitor cell for these lesions.

Adenocarcinoma↗

PubMed1988

Flow cytometric analysis of DNA ploidy pattern from deparaffinized formalin-fixed gastric cancer tissue.

Histologically processed tissue from gastric cancers has been analyzed by flow cytometry in an attempt to correlate DNA ploidy pattern and behavior of the tumor. Of the mucosal and submucosal cancers (so-called early, all stage I in the present series), 62.7% show a diploid DNA pattern and 37.3% show a single aneuploid pattern. Of the deeply infiltrating (beyond the submucosa) cancers (stage II and III), 52.1% are single aneuploid and 47.9% are multiploid. While stage-I patients are all alive at the end of the follow-up period (6 years), in stage II and III cases Cox's regression model shows that the hazard function depends on DNA pattern: survival is negatively influenced by multiploidy. On this basis, it may be assumed that the DNA pattern is a useful prognostic indicator of gastric cancer. As expected, in Cox's regression model an even more important negative correlation exists between survival and stage: single aneuploid cases in stage II have a better prognosis than those in stage III. Instead, no correlation is found between histological cancer subtype (Laurén and WHO classifications), grade and DNA pattern.

Adult↗

PubMed2026

Stage-Independent Real-Time Subtype Classification and Comprehensive Biopsy Profiling of Urothelial Carcinomas by the Lund Taxonomy System.

Bladder cancer is a heterogeneous malignancy with diverse clinical outcomes, and conventional pathological assessment alone is insufficient to capture its underlying biology. Gene expression profiling can stratify tumors into molecular subtypes with prognostic and predictive potential, but the reliability of transcriptomic classification and its clinical utility remains to be established. The translational/observational UROSCANSEQ study (ISRCTN15459149) prospectively evaluates RNA-based Lund Taxonomy (LundTax) molecular subtype classification in a clinical setting. Among 784 consecutive biopsies collected between 2018 and 2022, RNA sequencing was successful for 90% of all biopsies, encompassing 662 bladder cancer patients with a stage distribution of 48% Ta, 27% T1, 24% ≥T2, and 1% CIS. We demonstrate that the LundTax subtype classification algorithm, applied to individual samples, accurately identifies cancer cell phenotypes with characteristic gene and protein expression patterns in a manner robust to RNA quality, data preprocessing strategies, and batch effects, supporting its clinical feasibility across both non-muscle-invasive and muscle-invasive disease. We further extend the LundTax framework by incorporating single-sample molecular risk scores reflecting tumor grade, proliferation, and progression risk, as well as tumor microenvironment signatures. Both risk scores and overall immune and stromal content in biopsies were significantly associated with an increased risk of clinical progression in noninvasive disease. In a separate analysis of the relative cellular composition of the tumor microenvironment, however, only the fraction of natural killer cells remained significant. Together, the expanded LundTax system provides a comprehensive molecular portrait of individual tumor biopsies. By explicitly separating cancer cell-intrinsic phenotypes, prognostic indexes, and microenvironmental signals, the framework minimizes biological confounding and establishes a strong foundation for future studies evaluating clinical outcomes and treatment responses.

Humans↗

PubMed1984

The prognostic significance of histologic subtyping in small cell carcinoma of the lung.

Previously untreated patients with small cell carcinoma of the lung (SCCL), who were treated at the Medical College of Wisconsin with combined chemotherapy and radiation therapy, were retrospectively subtyped according to the 1981 World Health Organization Lung Cancer Classification. Of 54 evaluated patients, 27 (50%) had "oat cell" subtype, 22 (41%) "intermediate cell" variety, and five (9%) were classified as "combined" type. There was no significant difference in response to therapy or median survival between the subtypes. In addition to the absence of prognostic significance among the subtypes, there were many technical factors affecting accuracy of subtyping, including tissue-crushing artifacts, size of biopsy materials, fixation of tissue samples, and variation of subtypes within the same biopsy. We conclude that subtyping of SCCL should not be construed as a prognostic tool or guideline to therapy. However, the recognition that SCCL may manifest in a variety of histologic patterns, some of which may be misinterpreted as a histology other than SCCL, is probably more important for choice of therapy and prognosis than the individual subtypes.

Carcinoma, Small Cell↗

PubMed1988

Histopathologic classification of small cell lung cancer. Changing concepts and terminology.

Considerable attention has been devoted to the diagnosis of small cell lung carcinoma (SCLC) and its subtypes. In the literature contradictory opinions have been published concerning the clinical implications of subtyping, largely because of the different criteria used by different pathologists. This article is a consensus report by the Pathology Committee of the International Association for the Study of Lung Cancer. The following classification of SCLC is recommended: (1) Small cell carcinoma. This subtype includes most of the tumors previously included in the oat cell and intermediate subtypes. More than 90% of untreated SCLC fall into this category. (2) Mixed small cell/large cell carcinoma. This subtype, which may be associated with a poor prognosis and response to therapy, contains a spectrum of cell types ranging from typical SCLC to larger cells having prominent nucleoli and resembling large cell carcinoma. (3) Combined small cell carcinomas. Typical SCLC elements are intimately admixed with areas of differentiated squamous cell or adenocarcinoma. This simplified classification of SCLC will facilitate uniformity in the diagnosis and further our understanding of the clinical significance of the rarer SCLC with variant morphologies.

Carcinoma, Small Cell↗

PubMed2026

Hierarchical modeling of tumor subtypes in cell lines using large-scale genomic datasets.

Cancer cell lines (CLs) are widely used to study tumor biology and drug response, yet their translational relevance is often limited by inaccurate subtype annotations. Existing CL-tumor matching approaches are frequently constrained by flat classification schemes, weak subtype definitions, and the exclusion of normal tissue references, leading to potential confounding of tumor-specific and tissue-of-origin signals. To address these limitations, a hierarchical classification (HC) framework is presented in which CLs are aligned with patient tumors across biological resolutions, from organ to molecular subtype. Gene expression profiles from 802 CLs, 5,612 tumors from The Cancer Genome Atlas (TCGA) , and 8,939 non-cancerous tissues were integrated to separate oncogenic signals from tissue-specific signals. Node-specific features were selected using maximum relevance minimum redundancy, and balanced accuracies of 89% in cross-validation and 75%, and 80% on external datasets were achieved. Through the framework, 43 CLs were reassigned, and clinically relevant underrepresented subtypes were identified.

cancer cell lines↗

PubMed1987

A morphologic study of childhood lymphoma of the diffuse "histiocytic" type. The Pediatric Oncology Group experience.

Of 227 cases of pediatric non-Hodgkin's lymphoma with adequate histopathologic material for review, 72 (32%) were classified as diffuse "histiocytic" lymphoma (DHL). These cases were further divided into different morphologic subtypes according to the Lukes-Collins classification, and the National Cancer Institute Working Formulation, to ascertain whether there were any significant prognostic differences among the different subtypes. The results of our study showed that 40 patients were classified as immunoblastic lymphomas, and 32 were called large follicular center cell (FCC) tumors. Of the 40 patients with immunoblastic histology, 19 had morphologic features of the clear cell type and were interpreted as consistent with T-immunoblastic lymphomas; an additional two had polymorphous features also consistent with T-cell type: 17 had plasmacytoid features, and were morphologically classified as B-immunoblastic lymphomas; two could not be subtyped. Of the 32 patients with morphologic features of FCC lymphomas, 29 were classified as large noncleaved type, and three as large cleaved type. A clinicopathologic analysis showed that 90% of the patients obtained complete remission, and there were no significant differences in complete remission rate among the different morphologic subtypes of DHL. The estimated five year disease-free survival for all patients was over 70%, with no failure after the second year; and there were no significant differences in the disease-free survival among the different subtypes. The only clinical differences that we found, were that patients with lymphomas of FCC (large noncleaved) type were younger (P = 0.01); had less nodal involvement (P = 0.03); and had more organ involvement (P less than 0.01). We conclude that the morphologic subclassification of DHL in children currently has limited clinical prognostic significance.

Adolescent↗

PubMed1980

Influence of histologic subtype of small cell carcinoma of the lung on clinical presentation, response to therapy, and survival.

Patients with small cell carcinoma of the lung (SCCL) were histologically subtyped according to the Working Party for Therapy of Lung Cancer classification and were treated with combination chemotherapy. Of the 103 patients studied, 54 had the lymphocyte-like (oat cell) subtype, 41 had the intermediate cell subtype, and 8 had a mixture of the two. No significant difference in initial performance status, extent of disease, chemotherapeutic response rate, or survival (median, 10.2 mo) was noted among the histologic subtypes. When the histologic subtype of the primary biopsy tissue was compared with the subtype of other pathology specimens from the same patient, concordance of subtype was present in 74% of the patients. In the remaining 26%, two or three histologic subtypes were present. This study demonstrates no clinically significant differences among the various histologic subtypes of SCCL in patients extensively staged and treated with aggressive cytotoxic therapy. Because of this and because concurrent biopsy tissues from multiple sites in the same patient may vary in subtype, we conclude that prognostic or therapeutic decisions should not be based on SCCL subtype.

Carcinoma, Small Cell↗

PubMed2025

MO-GCAN: multi-omics integration based on graph convolutional and attention networks.

MOTIVATION: Cancer subtypes play a critical role in disease progression, prognosis, and treatment, making their detection essential for tailoring precision medicine. Studies have shown that multi-omics integration outperforms single-omics approaches in cancer subtyping tasks. However, due to the high-dimensionality of multi-omics data, many existing studies either fail to capture the correlation between true labels and learned features, or lack sufficient capacity to model complex biological representations. These limitations hinder the full potential of leveraging the rich and complementary information embedded in multi-omics datasets. RESULT: We propose a framework that leverages supervised feature learning and classification based on a graph-based learning approach with attention mechanism for cancer subtyping. More specifically, we train graph convolutional network models on each omics dataset to extract latent representations, which are then concatenated to form a comprehensive multi-omics feature embedding. We further develop sample fusion network based on the omics-specific graphs, incorporating the derived features and feeding them into a graph attention model for subtype classification. This two-stage multi-omics framework is applied to eight cancer types, with performance evaluated in terms of test accuracy, training time, macro-averaged precision, recall, and F-score. Experimental results show that the proposed method outperforms state-of-the-art approaches across various cancer types. Additionally, we provide empirical evidence supporting the hypothesis that retaining a limited number of high-confidence edges and utilizing enriched embeddings from intermediate graph neural network layers can improve predictive performance. AVAILABILITY AND IMPLEMENTATION: Data and the code are available at https://github.com/YD-00/MO-GCAN-Updated.git.

Neoplasms↗

PubMed2021

Automated Extraction of Tumor Staging and Diagnosis Information From Surgical Pathology Reports.

PURPOSE: Typically stored as unstructured notes, surgical pathology reports contain data elements valuable to cancer research that require labor-intensive manual extraction. Although studies have described natural language processing (NLP) of surgical pathology reports to automate information extraction, efforts have focused on specific cancer subtypes rather than across multiple oncologic domains. To address this gap, we developed and evaluated an NLP method to extract tumor staging and diagnosis information across multiple cancer subtypes. METHODS: The NLP pipeline was implemented on an open-source framework called Leo. We used a total of 555,681 surgical pathology reports of 329,076 patients to develop the pipeline and evaluated our approach on subsets of reports from patients with breast, prostate, colorectal, and randomly selected cancer subtypes. RESULTS: Averaged across all four cancer subtypes, the NLP pipeline achieved an accuracy of 1.00 for International Classification of Diseases, Tenth Revision codes, 0.89 for T staging, 0.90 for N staging, and 0.97 for M staging. It achieved an F1 score of 1.00 for International Classification of Diseases, Tenth Revision codes, 0.88 for T staging, 0.90 for N staging, and 0.24 for M staging. CONCLUSION: The NLP pipeline was developed to extract tumor staging and diagnosis information across multiple cancer subtypes to support the research enterprise in our institution. Although it was not possible to demonstrate generalizability of our NLP pipeline to other institutions, other institutions may find value in adopting a similar NLP approach-and reusing code available at GitHub-to support the oncology research enterprise with elements extracted from surgical pathology reports.

Humans↗

PubMed2026

Integrating molecular subtypes, genomics and functional dependencies to identify context-specific therapeutic vulnerabilities in small cell lung cancer.

Small cell lung cancer is one of the most aggressive malignancies, characterized by rapid tumor growth, early metastatic spread and extremely poor survival. Although most patients initially respond to platinum-based chemotherapy, relapse is almost inevitable and treatment options at recurrence remain limited. The recent introduction of immune checkpoint inhibitors has provided only modest clinical benefit, largely due to the fact that these tumors are immunologically cold. These limitations highlight the urgent need to better understand the molecular features of small cell lung cancer in order to identify more effective therapeutic strategies. In this review, we summarize current knowledge of the molecular landscape of small cell lung cancer, with particular emphasis on transcriptome-based classifications that have identified four major molecular subtypes defined by distinct transcriptional regulators and gene expression programs. We discuss how these classifications have improved the biological understanding of the disease and stimulated efforts to develop subtype-specific therapeutic strategies. At the same time, we highlight important limitations of this framework, including the remarkable transcriptional plasticity of tumor cells, which allows dynamic transitions between subtypes and may contribute to therapeutic resistance. To address these challenges, we examine additional molecular features that may represent more stable vulnerabilities, including recurrent genomic alterations, such as the widespread loss of tumor suppressor genes or oncogene amplifications through extrachromosomal DNA. We also discuss emerging approaches aimed at identifying novel context-specific cancer dependencies, including genome-scale functional screens in vitro and in vivo and genetic restraint analyses. Finally, we consider the growing potential of liquid biopsy strategies, which exploit the high level of circulating tumor DNA in patients with this disease to detect clinically relevant genomic alterations and monitor tumor evolution. Overall, this review highlights both the opportunities and challenges associated with molecular stratification in small cell lung cancer. The integration of transcriptional classifications with genomic and functional approaches may help identify more robust therapeutic vulnerabilities and guide the development of more effective treatments for this highly aggressive disease.

Cancer vulnerabilities↗

PubMed2025

MWENA: a novel sample re-weighting-based algorithm for disease classification and data interpretation using extracellular vesicles omics data.

BACKGROUND AND OBJECTIVE: Extracellular vesicles (EVs), considered as a form of liquid biopsy, have gained significant attention in recent years due to their stability and the preservation of disease markers. Research studies underscore the clinical significance of molecules found in EVs, highlighting their role as communicative mediators between cells. However, analyzing this data is challenging due to noisy measurements, having far more variables than samples, and some groups (e.g., disease subtypes or experimental conditions) having much less data than others. We therefore develop an algorithm to address aforementioned challenges for the classification of imbalanced EVs omics data. METHODS AND RESULTS: We propose the EV Meta-Weight Elastic Net Algorithm (MWENA), which utilizes logistic regression with elastic net regularization for the classification and identification of EV signatures, effectively addressing the challenges posed by high-dimensional small sample sizes. To mitigate issues related to class imbalance and high noise levels, MWENA incorporates an automatic sample re-weighting function, which uses a meta-net to adaptively learn generalizable patterns directly from the data itself. We validate the MWENA algorithm on both simulated data and EVs omics data, covering six classification tasks that involve four different types of diseases (pancreatic ductal adenocarcinoma, interstitial lung diseases, colorectal cancer, and ovarian cancer) and three clinical scenarios (disease diagnosis, disease-stage screening, and disease-subtype classification). Compared to other machine learning methods, MWENA demonstrates superiority in identifying small class samples and achieves the highest scores in both sensitivity and G-means. Biological analysis is also performed to further explore the significance of selected signatures as biological markers and their roles in disease mechanisms. CONCLUSIONS: We anticipate that our proposed approach will take a modest step in harnessing EV omics data to discover biomarkers, aiding researchers in gaining a comprehensive understanding of biological processes.

Extracellular Vesicles↗

PubMed2026

Immune subtyping of colorectal adenoma identifies a subtype with activated adaptive immunity ahead of progressing to cancer.

BACKGROUND: Colorectal adenomas (CRA) represent precursor lesions with varying risks of malignant transformation. However, molecular subtyping, particularly immune-related classification, remains underexplored in adenomas. This study aims to characterize the immune landscape of CRA through immune subtyping and evaluate its association with cancer progression, gene expression signatures, and functional pathways. METHODS: We conducted a retrospective analysis of transcriptomic data from multiple cohorts of CRA samples. Immune subtypes were identified using non-negative matrix factorization (NMF) based on immune-related genes. Diverse deconvolution algorithms were used to estimate immune cell infiltration. The immune status alteration in premalignant lesion was further consolidated by single-cell transcriptome data. Differential gene expression analysis was performed between subtypes, followed by functional enrichment analyses (Gene Ontology [GO] and Kyoto Encyclopedia of Genes and Genomes [KEGG]). RESULTS: Two distinct immune subtypes were identified: an immune-enriched subtype characterized by high lymphocyte infiltration and elevated expression of immune-related genes, and an immune-deficient subtype with suppressed immune activity. Differential expression analysis revealed significant upregulation of immune response genes (e.g., CD4, CD86, HLA-DRA) in the immune-enriched subtype. GO and KEGG analyses highlighted enrichments in leukocyte transendothelial migration, chemokine signaling, and antigen processing and presentation pathways. Single-cell result revealed an early occurrence of TIGIT activation and exhausted CD8 T cell features in adenoma when compared to normal tissue. CONCLUSION: This study delineates distinct immune subtypes within CRAs. The immune-enriched subtype demonstrates activated adaptive immunity and may reflect a higher potential for immune surveillance, while the immune-deficient subtype exhibits stromal features suggestive of progressive transformation. These findings provide insights into early immune microenvironment alterations and may inform strategies for risk stratification and immunoprevention in colorectal carcinogenesis.

Colorectal adenoma↗

PubMed2026

Expression patterns of potential targets for antibody-directed therapy in metastatic castration-resistant prostate cancer patients.

INTRODUCTION: Survival in metastatic castration-resistant prostate cancer (mCRPC) patients remains limited and treatment is complicated by tumor heterogeneity. As antibody-based therapeutics emerge, identifying actionable antigen targets and patient subgroups most likely to benefit is essential. MATERIALS & METHODS: Gene expression of 62 antibody-targetable proteins was analyzed in 296 mCRPC biopsies. These genes encode proteins targeted by approved or investigational antibody-based cancer therapeutics. Associations between target expression with genomic classifications and transcriptomic subtypes were evaluated. Target expression was also assessed in tumors with low expression of established mCRPC targets. Subgroup-specific targets were validated in an independent cohort and single-cell transcriptomics. RESULTS: Established targets KLK2, FOLH1 (PSMA) and STEAP1 showed the highest median expression across the cohort. Target expression did not correlate with genomic classifications, including homologous recombination deficiency, microsatellite instability, CDK12, TP53, PTEN or AR alterations Target expression did associate with transcriptomic subtypes: CRPC-AR (driven by androgen receptor-signaling) and CRPC-SCL (stem cell-like features, AP-1/YAP/TAZ-driven), displayed the highest expression of multiple targets, including KLK2, FOLH1, and SLC44A4. CRPC-NE (neuroendocrine phenotype) showed heterogeneous expression, with high CD46 expression, whereas CRPC-WNT (Wnt-signaling driven) generally showed low target expression. Notably, CD46 was highly expressed in tumors with low KLK2, FOLH1, and STEAP1 expression, a subgroup associated with poor prognosis. CONCLUSIONS: Although several antibody targets showed broad expression in mCRPC-tumors, expression varied by transcriptomic subtype. Subgroups such as CRPC-WNT expressed fewer targets, suggesting the need for alternative therapeutic strategies. CD46 emerged as a promising target, with wide expression across multiple subtypes, including clinically challenging CRPC-NE and mCRPC tumors lacking expression of established targets.

Humans↗

PubMed2025

HallmarkGraph: a cancer hallmark informed graph neural network for classifying hierarchical tumor subtypes.

MOTIVATION: Accurate tumor subtype diagnosis is crucial for precision oncology, yet current methodologies face significant challenges. These include balancing model accuracy with interpretability and the high costs of generating multi-omics data in clinical settings. Moreover, there is a lack of validated models capable of classifying hierarchical tumor subtypes across a comprehensive pan-cancer cohort. RESULTS: We present a graph neural network, HallmarkGraph, the first biologically informed model developed to classify hierarchical tumor subtypes in human cancer. Inspired by cancer hallmarks, the model's architecture integrates transcriptome profiles and gene regulatory interactions to perform multi-label classification. We evaluate the model on a comprehensive pan-cancer cohort comprising 11 476 samples from 26 primary cancers with 405 subtypes up to eight levels. The model demonstrates exceptional performance, achieving 5-fold cross-validation accuracy between 85% and 99% for tumor subtypes labeled with increasing details of genomic information. It also shows good generalizability on a validation dataset of 887 samples, assessed using three metrics that consider tumor subtypes at individual, combined, and sample levels. Benchmarking and ablation experiments show that hallmark-based embeddings slightly influence model performance, while the integrated multilayer perceptron plays a significant role in determining classifier accuracy. Additionally, we use the SHAP method to link cancer hallmarks with genes, identifying key features that influence model decisions. Our findings present a biologically informed machine learning framework capable of tracking tumor transcriptomic trajectories and distinguishing inter- and intra-tumor heterogeneity in pan-cancer. This approach holds promise for enhancing cancer diagnostics. AVAILABILITY AND IMPLEMENTATION: HallmarkGraph is accessible at https://github.com/laixn/HallmarkGraph.

Humans↗

PubMed1994

[An extended operation for scirrhous gastric cancer--its significance and procedure].

We have been performing an extended operation for scirrhous gastric cancer, namely left upper abdominal exenteration plus Appleby's method, and compared the prognosis among macroscopic subtype classifications. The surgical procedure of left upper abdominal exenteration plus Appleby's method is to resect the neighboring organs of the stomach in an en bloc fashion. It consists of total gastrectomy, partial pancreatectomy, splenectomy, transverse colectomy, cholecystectomy, resection of the left adrenal gland, and ligation and resection of the common hepatic artery. D4 lymph node dissection is carried out. Extended operations were performed for 45 cases of P0-1 scirrhous gastric cancer in 1983-1990. There were 25, 17 and 3 cases of giant fold type, erosive type and stenotic type scirrhous gastric cancers, respectively. Positive lymph node metastases were found in 64% of giant fold type cases and 82% of erosive type cases. In Po, T3-4 cases, cytology of intraoperative peritoneal lavage was positive in 27% of giant fold type cases and 50% of erosive type cases. Although there were no obvious macroscopic invasions found during the operative procedure, later histological study showed invasion to the left adrenal gland in 14% of giant fold type cases and 0% of erosive type cases, and invasion to the transverse colon in 21% of giant fold type cases and 15% of fold type and erosive type scirrhous gastric cancer according to the histological invasion of transverse colon were 25% in positive cases and 27% in negative cases (p = 0.27), 0% in positive cases and 27% in negative cases (p = 0.27), respectively. In giant fold type scirrhous gastric cancer, the extended operation is favorable. In erosive type scirrhous gastric cancer, peritoneal dissemination is observed with a high incidence. Therefore, prophylactic therapy is significant.

Adenocarcinoma, Scirrhous↗