Methylation profiling and machine learning distinguishes primary lung squamous cell carcinomas from head and neck metastases
Ontology highlight
ABSTRACT:
PROVIDER: PRJNA510619 | ENA |
REPOSITORIES: ENA
Similar Datasets
Project description:Head and neck squamous cell carcinoma (HNSC) patients are at risk of suffering from both pulmonary metastases or a second squamous cell carcinoma of the lung (LUSC). Differentiating pulmonary metastases from primary lung cancers is of high clinical importance, but not possible in most cases with current diagnostics. To address this, we performed DNA methylation profiling of primary tumors and trained three different machine learning methods to distinguish metastatic HNSC from primary LUSC. We developed an artificial neural network that correctly classi- fied 96.4% of the cases in a validation cohort of 279 patients with HNSC and LUSC as well as normal lung controls, outperforming support vector machines (95.7%) and random forests (87.8%). Prediction accuracies of more than 99% were achieved for 92.1% (neural network), 90% (support vector machine), and 43% (random forest) of these cases by applying thresholds to the resulting probability scores and excluding samples with low confidence. As independent clinical validation of the approach, we analyzed a series of 51 patients with a history of HNSC and a second lung tumor, demonstrating the correct classifications based on clinicopathological properties. In summary, our approach may facilitate the reliable diagnostic differentiation of pulmonary metastases of HNSC from primary LUSC to guide therapeutic decisions.
2019-09-10 | GSE124052 | GEO
Project description:Objectives Our goal was to evaluate the diagnostic value of DNA methylation analysis in combination with machine learning to differentiate pleural mesothelioma (PM) from important histopathological mimics. Material and methods DNA methylation data of PM, lung adenocarcinomas, lung squamous cell carcinomas and chronic pleuritis was used to train a random forest as well as a support vector machine. These classifiers were validated using an independent validation cohort including pleural carcinosis and pleomorphic variants of lung adeno- and squamous cell carcinomas. Furthermore, we used a deconvolution method to estimate the composition of the tumor microenvironment. Results T-distributed stochastic neighbor embedding clearly separated PM from lung adenocarcinomas and squamous cell carcinomas, but there was a considerable overlap between chronic pleuritis specimens and PM with low tumor cell content. While both machine learning algorithms achieved comparable accuracies in a nested cross validation on the training cohort (random forest: 94.9%; support vector machine: 95.5%), the support vector machine outperformed the random forest in distinguishing PM from chronic pleuritis. Differential methylation analysis revealed promoter hypermethylation in PM specimens, including the tumor suppressor genes BCL11B, EBF1, FOXA1, and WNK2. Furthermore, we observed comparable accuracies for the support vector machine on the validation cohort (97.1%) while the random forest performed considerably worse (89.9%). Deconvolution of the stromal and immune cell composition revealed higher rates of regulatory T-cells and endothelial cells in tumor specimens and a heterogenous inflammation including macrophages, B-cells and natural killer cells in chronic pleuritis. Conclusion DNA methylation in combination with machine learning is a promising tool to reliably differentiate PM from chronic pleuritis and lung cancer, including pleomorphic carcinomas. Furthermore, our study highlights new candidate genes for PM carcinogenesis and shows that deconvolution of DNA methylation data can provide reasonable insights into the composition of the tumor microenvironment.
2022-05-17 | GSE203061 | GEO
Project description:In head and neck squamous cell cancers (HNSCs) that present as metastases with an unknown primary (HNSC-CUPs), the identification of a primary tumor improves therapy options and increases patient survival. However, the currently available diagnostic methods are laborious and do not offer a sufficient detection rate. Predictive machine learning models based on DNA methylation profiles have recently emerged as a promising technique for tumor classification. We applied this technique to HNSC to develop a tool that can improve the diagnostic workup for HNSC-CUPs. On a reference cohort of 405 primary HNSC samples, we developed four classifiers based on different machine learning models (random forest (RF), neural network (NN), elastic net penalized logistic regression (LOGREG), support vector machine (SVM)) that predict the primary site of HNSC tumors from their DNA methylation profile. The classifiers achieved high classification accuracies (RF=83%, NN=88%, LOGREG=SVM=89%) on an independent cohort of 64 HNSC metastases. Further, the NN, LOGREG, and SVM models significantly outperformed p16 status as a marker for an origin in the oropharynx. In conclusion, the DNA methylation profiles of HNSC metastases are characteristic for their primary sites and the classifiers developed in this study, which are made available to the scientific community, can provide valuable information to guide the diagnostic workup of HNSC-CUP.
2021-12-08 | GSE171994 | GEO
Project description:DNA methylation-based machine learning classification distinguishes pleural mesothelioma from chronic pleuritis, pleural carcinosis, and pleomorphic lung carcinomas
| PRJNA838375 | ENA
Project description:This study was designed to find patterns of gene expression in colon cancer samples that correlate with clinical stage. Ten samples were profiled from each of the following sample types: normal colon tissue, noncancerous adenomas, Dukes B carcinomas, Dukes C carcinomas, Dukes D carcinomas, and liver metastases. See also E-TIGR-14, E-TIGR-16
2005-01-01 | E-TIGR-15 | biostudies-arrayexpress
Project description:This study was designed to find patterns of gene expression in colon cancer samples that correlate with clinical stage. Ten samples were profiled from each of the following sample types: normal colon tissue, noncancerous adenomas, Dukes B carcinomas, Dukes C carcinomas, Dukes D carcinomas, and liver metastases. See also E-TIGR-14, E-TIGR-15
2005-01-01 | E-TIGR-16 | biostudies-arrayexpress
Project description:Gene expression profiles were generated from 199 primary breast cancer patients. Samples 1-176 were used in another study, GEO Series GSE22820, and form the training data set in this study. Sample numbers 200-222 form a validation set. This data is used to model a machine learning classifier for Estrogen Receptor Status. RNA was isolated from 199 primary breast cancer patients. A machine learning classifier was built to predict ER status using only three gene features.
2013-01-01 | E-GEOD-29210 | biostudies-arrayexpress
Project description:Machine Learning Models Predict the Primary Sites of Head and Neck Squamous Cell Carcinoma Metastases Based on DNA Methylation
| PRJNA721669 | ENA
Project description:This study was designed to find patterns of gene expression in colon cancer samples that correlate with clinical stage. Ten samples were profiled from each of the following sample types: normal colon tissue, noncancerous adenomas, Dukes B carcinomas, Dukes C carcinomas, Dukes D carcinomas, and liver metastases. See also E-TIGR-14, E-TIGR-15
2005-01-01 | E-TIGR-14 | biostudies-arrayexpress
Project description:[original title] Metastatic Canine Mammary Carcinomas can be Identified by a Gene Expression Profile that partly Overlaps with Human Breast Cancer Profiles. Introduction: Similar to human breast cancer mammary tumors of the female dog are commonly associated with a fatal outcome due to the development of distant metastases. However, the molecular defects leading to metastasis are largely unknown and the value of canine mammary carcinoma as a model for human breast cancer is unclear. In this study, we analyzed the gene expression signatures associated with mammary tumor metastasis and asked for parallels with the human equivalent. Methods: Messenger RNA expression profiles of twenty-seven lymph node metastasis positive or negative canine mammary carcinomas were established by microarray analysis. Differentially expressed genes were functionally characterized and associated with molecular pathways. The findings were also correlated with published data on human breast cancer. Results: Metastatic canine mammary carcinomas had 1,011 significantly differentially expressed genes when compared to non-metastatic carcinomas. Metastatic carcinomas had a significant up-regulation of genes associated with cell cycle regulation, matrix modulation, protein folding and proteasomal degradation whereas cell differentiation genes, growth factor pathway genes and regulators of actin organization were significantly down-regulated. Interestingly, 265 of the 1,011 differentially expressed canine genes are also related to human breast cancer and, vice versa, parts of a human prognostic gene signature were identified in the expression profiles of the metastatic canine tumors. Conclusions: Metastatic canine mammary carcinomas can be discriminated from non-metastatic carcinomas by their gene expression profiles. More than one third of the differentially expressed genes are also described of relevance for human breast cancer. Many of the differentially expressed genes are linked to functions and pathways which appear to be relevant for the induction and maintenance of metastatic progression and may represent new therapeutic targets. Furthermore, dogs are in some aspects suitable as a translational model for human breast tumors in order to identify prognostic molecular signatures and potential therapeutic targets. Thirteen simple mammary carcinomas with invasive growth and lymph node metastases at the time of tumor resection and 14 simple carcinomas without lymph node metastases were included in the study. None of the patients had radiographically detectable pulmonary metastases at the time of tumor resection. Distant metastases as the cause of death were determined postoperatively by radiographic detection of metastases or necropsy. Selection criteria for carcinomas without lymph node metastases included an invasive growth, a negative lymph node status, a histological grade III and a minimal tumor diameter above the average of the lymph node positive tumors (> 2.42 cm). All animals with non-metastatic carcinomas had a survival rate of over 24 months except animal no. 22 (2627) which developed radiographic detectable lung metastases 8 months after surgery. Tumor and lymph node histologies were evaluated independently by two board-certified pathologists, following the criteria of the WHO classification of canine mammary tumors and the Nottingham grading system. All 27 tumors were simple carcinomas and characterized by an invasive, mostly solid growth pattern, marked cellular pleomorphism, anisokaryosis and 3 or more mitotic figures per high power field.
2010-12-01 | E-GEOD-20718 | biostudies-arrayexpress