Predictor Of Survival In Cancer Research Articles

Breast cancer survival prediction using an automated mitosis detection pipeline.

Mitotic count (MC) is the most common measure to assess tumor proliferation in breast cancer patients and is highly predictive of patient outcomes. It is, however, subject to inter- and intraobserver variation and reproducibility challenges that may hamper its clinical utility. In past studies, artificial intelligence (AI)-supported MC has been shown to correlate well with traditional MC on glass slides. Considering the potential of AI to improve reproducibility of MC between pathologists, we undertook the next validation step by evaluating the prognostic value of a fully automatic method to detect and count mitoses on whole slide images using a deep learning model. The model was developed in the context of the Mitosis Domain Generalization Challenge 2021 (MIDOG21) grand challenge and was expanded by a novel automatic area selector method to find the optimal mitotic hotspot and calculate the MC per 2 mm2. We employed this method on a breast cancer cohort with long-term follow-up from the University Medical Centre Utrecht (N = 912) and compared predictive values for overall survival of AI-based MC and light-microscopic MC, previously assessed during routine diagnostics. The MIDOG21 model was prognostically comparable to the original MC from the pathology report in uni- and multivariate survival analysis. In conclusion, a fully automated MC AI algorithm was validated in a large cohort of breast cancer with regard to retained prognostic value compared with traditional light-microscopic MC.

Enhanced ovarian cancer survival prediction using temporal analysis and graph neural networks

Ovarian cancer is a formidable health challenge that demands accurate and timely survival predictions to guide clinical interventions. Existing methods, while commendable, suffer from limitations in harnessing the temporal evolution of patient data and capturing intricate interdependencies among different data elements. In this paper, we present a novel methodology which combines Temporal Analysis and Graph Neural Networks (GNNs) to significantly enhance ovarian cancer survival rate predictions. The shortcomings of current processes originate from their disability to correctly seize the complex interactions amongst diverse scientific information units in addition to the dynamic modifications that arise in a affected person`s nation over time. By combining temporal information evaluation and GNNs, our cautioned approach overcomes those drawbacks and, whilst as compared to preceding methods, yields a noteworthy 8.3% benefit in precision, 4.9% more accuracy, 5.5% more advantageous recall, and a considerable 2.9% reduction in prediction latency. Our method’s Temporal Analysis factor uses longitudinal affected person information to perceive good-sized styles and tendencies that offer precious insights into the direction of ovarian cancer. Through the combination of GNNs, we offer a robust framework able to shoot complicated interactions among exclusive capabilities of scientific data, permitting the version to realize diffused dependencies that would affect survival results. Our paintings have tremendous implications for scientific practice. Prompt and correct estimation of the survival price of ovarian most cancers allows scientific experts to customize remedy regimens, manipulate assets efficiently, and provide individualized care to patients. Additionally, the interpretability of our version`s predictions promotes a collaborative method for affected person care via way of means of strengthening agreement among scientific employees and the AI-driven selection help system. The proposed approach not only outperforms existing methods but also has the possible to develop ovarian cancer treatment by providing clinicians through a reliable tool for informed decision-making. Through a fusion of Temporal Analysis and Graph Neural Networks, we conduit the gap among data-driven insights and clinical practice, proposing a capable opportunity for refining patient outcomes in ovarian cancer management operations.

Machine learning applications in breast cancer survival and therapeutic outcome prediction based on multi-omic analysis.

The high heterogeneity within and between breast cancer patients complicates treatment determination and prognosis assessment. Treatment decision-making is influenced by various factors, such as tumor subtype, histological grade, and genotype, necessitating personalized treatment strategies. Prognostic outcomes vary significantly depending on patient-specific conditions. As a critical branch of artificial intelligence, machine learning efficiently handles large datasets and automates decision-making processes. The introduction of machine learning offers new solutions for breast cancer treatment selection and prognosis assessment. In the field of cancer therapy, traditional methods for predicting treatment and survival outcomes often rely on single or few biomarkers, limiting their ability to capture the complexity of biological processes comprehensively. Machine learning analyzes patients' multi-omic data and the intricate patterns of variations during cancer initiation and progression to predict patients' survival and treatment outcomes. Consequently, it facilitates the selection of appropriate therapeutic interventions to implement early intervention and improve treatment efficacy for patients. Here, we first introduce common machine learning methods, and then elaborate on the application of machine learning in the field of survival prediction and prognosis from two aspects: evaluating survival and predicting treatment outcomes for breast cancer patients. The aim is to provide breast cancer patients with precise treatment strategies to improve therapeutic outcomes and quality of life.

Enhancing Non-Small Cell Lung Cancer Survival Prediction through Multi-Omics Integration Using Graph Attention Network.

Background: Cancer survival prediction is vital in improving patients' prospects and recommending therapies. Understanding the molecular behavior of cancer can be enhanced through the integration of multi-omics data, including mRNA, miRNA, and DNA methylation data. In light of these multi-omics data, we proposed a graph attention network (GAT) model in this study to predict the survival of non-small cell lung cancer (NSCLC). Methods: The different omics data were obtained from The Cancer Genome Atlas (TCGA) and preprocessed and combined into a single dataset using the sample ID. We used the chi-square test to select the most significant features to be used in our model. We used the synthetic minority oversampling technique (SMOTE) to balance the dataset and the concordance index (C-index) to measure the performance of our model on different combinations of omics data. Results: Our model demonstrated superior performance, with the highest value of the C-index obtained when we used both mRNA and miRNA data. This demonstrates that the multi-omics approach could be effective in predicting survival. Further pathway analysis conducted with KEGG showed that our GAT model provided high weights to the features that are associated with the viral entry pathways, such as the Epstein-Barr virus and Influenza A pathways, which are involved in lung cancer development. From our findings, it can be observed that the proposed GAT model leads to a significantly improved prediction of survival by exploiting the strengths of multiple omics datasets and the findings from the enriched pathways. Our GAT model outperforms other state-of-the-art methods that are used for NSCLC prediction. Conclusions: In this study, we developed a new model for the survival prediction of NSCLC using the GAT based on multi-omics data. Our model showed outstanding predictive values, and the KEGG analysis of the selected significant features showed that they were implicated in pivotal biological processes underlying pathways such as Influenza A and the Epstein-Barr virus infection, which are linked to lung cancer progression.

Improve clinical feature-based bladder cancer survival prediction models through integration with gene expression profiles and machine learning techniques

BackgroundBladder cancer (BCa), one of the most common cancers worldwide, is characterized by high rates of recurrence, progression, and mortality. Machine learning algorithms offer promising advancements in enhancing predictive models. This study aims to develop robust machine learning models for predicting BCa survival using clinical and gene expression data. MethodsClinical data from BCa patients were obtained from the Surveillance, Epidemiology, and End Results database. Cox proportional hazards regression models assessed the association between clinical variables and overall survival. Machine learning algorithms, including logistic regression, random forest, XGBoost, decision tree, and LightGBM, were employed to predict survival at 1, 3, and 5 years. The TAGO database, combined with the data from The Cancer Genome Atlas and four databases from the Gene Expression Omnibus, which have available genomic data and clinical data, were selected. Gene expression data were transformed into gene sets data, and the performance of models based on clinical data and gene sets data and their combination were compared. Furthermore, the impact of model-derived scores on overall survival was evaluated. ResultsAmong 138,741 BCa patients with available clinical data, key independent predictors of survival included age, race, marital status, surgery, chemotherapy, radiation, and TNM stages. Clinical data machine learning (CML) models used these clinical predictors to achieve AUC values of 0.860, 0.821, and 0.804 in the testing sets for predicting survival at 1, 3, and 5 years, respectively. In the TAGO database, which has 863 patients with clinical and genomic data, the integrated clinical and gene expression machine learning model (IML) outperformed the CML and gene expression machine learning (GML) models in survival prediction. Patients with higher IML and GML model scores exhibited poorer survival outcomes. ConclusionsThis study successfully identifies key clinical and genomic predictors, a significant step forward in BCa research. The development of predictive models for BCa survival underscores the potential of integrated data approaches in improving BCa management and treatment strategies.

Overall survival prediction of gastric cancer using the gene signature of CT-detected extramural venous invasion combined with M2 macrophages infiltration

BackgroundCT-detected Extramural venous invasion (EMVI) is known as an independent risk factor for distant metastasis in patients with advanced gastric cancer (GC). However, the molecular basis is not clear. In colorectal cancer, M2 macrophages plays a vital role in determining EMVI. This study aimed to investigate the relationship between CT-detected EMVI and the M2 macrophages as well as prognosis predictionusing a radiogenomic approach.MethodWe utilized EMVI-related genes (from mRNA sequencing of 13 GC samples correlated with EMVI score by spearman analysis, P < 0.01) to overlap the co-expression genes of WGCNA module and M2 macrophages related genes (from mRNA data of 371 GC patients in TCGA database), generating a total of 136 genes. An EMVI-M2-prognosis-related hub gene signature was constructed by COX and least absolute shrinkage and selection operator (LASSO) analysis from a training cohort TCGA database (n = 371) and validated it in a validation cohort from GEO database (n = 357). High- and low-risk groups were divided by hub gene (EGFLAM and GNG11) signature-derived risk scores. We assessed its predictive ability through Kaplan-Meier (K-M) curve and COX analysis. Furthermore, we utilized ESTIMATE to detect tumor mutation burden (TMB) and evaluate sensitivity to immune checkpoint inhibitors (ICIs). Expression of hub genes was tested using western blotting and immunohistochemistry (IHC) analysis.ResultsThe overall survival (OS) was significantly reduced in the high-risk group (Training/Validation: AUC = 0.701/0.620; P < 0.001/0.003). Furthermore, the risk score was identified as an independent predictor of OS in multivariate COX regression analyses (Training/Validation: HR = 1.909/1.928; 95% CI: 1.225–2.974/1.308–2.844). The low-risk group exhibited significantly higher TMB levels (P = 1.6e− 07) and greater sensitivity to ICIs. Significant higher expression of hub-genes was identified on multiple GC cell lines and original samples. Hub-genes knockdown in gastric cancer cell lines inhibited their proliferation, metastatic and invasive capacity to varying degrees. In vivo experiments indicate that EGFLAM, as one of the hub genes, its high expression can serve as a biomarker for low response to immunotherapy.ConclusionOur study demonstrated EMVI-M2 gene signature could effectively predict the prognosis of GC tissue, reflecting the relationship between EMVI and M2 macrophages.

Multi-task deep latent spaces for cancer survival and drug sensitivity prediction.

Cancer is a very heterogeneous disease that can be difficult to treat without addressing the specific mechanisms driving tumour progression in a given patient. High-throughput screening and sequencing data from cancer cell-lines has driven many developments in drug development, however, there are important aspects crucial to precision medicine that are often overlooked, namely the inherent differences between tumours in patients and the cell-lines used to model them in vitro. Recent developments in transfer learning methods for patient and cell-line data have shown progress in translating results from cell-lines to individual patients in silico. However, transfer learning can be forceful and there is a risk that clinically relevant patterns in the omics profiles of patients are lost in the process. We present MODAE, a novel deep learning algorithm to integrate omics profiles from cell-lines and patients for the purposes of exploring precision medicine opportunities. MODAE implements patient survival prediction as an additional task in a drug-sensitivity transfer learning schema and aims to balance autoencoding, domain adaptation, drug-sensitivity prediction, and survival prediction objectives in order to better preserve the heterogeneity between patients that is relevant to survival. While burdened with these additional tasks, MODAE performed on par with baseline survival models, but struggled in the drug-sensitivity prediction task. Nevertheless, these preliminary results were promising and show that MODAE provides a novel AI-based method for prioritizing drug treatments for high-risk patients. https://github.com/UEFBiomedicalInformaticsLab/MODAE.

MSFN: a multi-omics stacked fusion network for breast cancer survival prediction.

Introduction: Developing effective breast cancer survival prediction models is critical to breast cancer prognosis. With the widespread use of next-generation sequencing technologies, numerous studies have focused on survival prediction. However, previous methods predominantly relied on single-omics data, and survival prediction using multi-omics data remains a significant challenge. Methods: In this study, considering the similarity of patients and the relevance of multi-omics data, we propose a novel multi-omics stacked fusion network (MSFN) based on a stacking strategy to predict the survival of breast cancer patients. MSFN first constructs a patient similarity network (PSN) and employs a residual graph neural network (ResGCN) to obtain correlative prognostic information from PSN. Simultaneously, it employs convolutional neural networks (CNNs) to obtain specificity prognostic information from multi-omics data. Finally, MSFN stacks the prognostic information from these networks and feeds into AdaboostRF for survival prediction. Results: Experiments results demonstrated that our method outperformed several state-of-the-art methods, and biologically validated by Kaplan-Meier and t-SNE.

Use of PROMISE criteria on PSMA-PET for prediction of overall survival in prostate cancer

Use of PROMISE criteria on PSMA-PET for prediction of overall survival in prostate cancer

Web-based prognostic tools for oral tongue cancer: An analysis of online predictors.

Oral tongue squamous cell carcinoma (OTSCC) often presents with aggressive clinical behaviour that may require multimodality treatment based on reliable prognostication. We aimed to evaluate the prognostic ability of five online web-based tools to predict the clinical behaviour of OTSCC resection and biopsy samples. A total of 135 OTSCC resection cases and 33 OTSCC biopsies were included to predict recurrence and survival. Area under the receiver operating characteristic curves (AUC), χ2 tests, and calibration plots constructed to estimate the prognostic power of each tool. The tool entitled 'Prediction of risk of Locoregional Recurrences in Early OTSCC' presented an accuracy of 82%. The tool, 'Head & Neck Cancer Outcome Calculator' for 10-year cancer-related mortality had an accuracy 77% and AUC 0.858. The other tool entitled 'Cancer Survival Rates' for 5-year mortality showed an accuracy of 74% and AUC of 0.723. For biopsy samples, 'Cancer Survival Prediction Calculators' predicted the recurrence free survival with an accuracy of 70%. Web-based tools can aid in clinical decision making of OTSCC. Three of five online web-based tools could predict recurrence risk and cancer-related mortality in resected OTSCC and one tool could help in clinical decision making for biopsy samples.

Biomonitoring Study of Toxic Metal(loid)s: Levels in Lung Adenocarcinoma Patients.

Lung cancer is a leading cause of cancer deaths worldwide. The aim of this study was to investigate heavy metal(loid)s (Cd, Pb, Hg, Cr, Mn, Mo, Ni, and As) in lung cancer patients in order to elucidate their role as lung cancer environmental risk factors. Sixty-three patients of both sexes with adenocarcinoma stage IIIB or IV were enrolled in this research. The heavy metal(loid) urine concentrations were measured using ICP-MS. Arsenic was quantified above 10 μg/L in 44.44% of the samples. Nickel urinary concentrations above the ToxGuide reference levels were found in 50.79% of the samples, while lead was quantified in 9.52% of the urine samples. The urinary chromium levels were above the mean ToxGuide levels in 41.27% of the patients and were significantly higher in men in comparison with women (p = 0.035). The chromium urinary concentrations were positively associated with the CRP serum levels (p = 0.037). Cadmium was quantified in 61.90% of the samples with levels significantly higher in females than in males (p = 0.023), which was associated with smoking habits. Mercury was measured above the limit of quantification in 63.49% of the samples and was not associated with amalgam dental fillings. However, the Hg urinary concentrations were correlated positively with the ALT (p = 0.02), AST (p < 0.001), and GGT (p < 0.001) serum levels. In 46.03% of the samples, the Mo concentrations were above 32 μg/L, the mean value for healthy adults according to the ToxGuide, and 9.52% of the patients had Mn levels higher than 8 μg/L, the reference value for healthy adults based on ToxGuide data. The obtained results are preliminary, and further studies are needed to have a deeper insight into metal(loid) exposure's association with lung cancer development, progression, and survival prediction.

The survival prediction of advanced colorectal cancer received neoadjuvant therapy—a study of SEER database

PurposeThe aim of study was to screen factors associated with the overall survival of colorectal cancer patients with lymph nodes metastasis who received neoadjuvant therapy and construct a nomogram model.MethodsAll enrolled subjects of the SEER database were randomly assigned to the training and testing group in a ratio of 3:2. The patients of Tangdu Hospital were seemed as validation group. Univariate cox regression analysis, lasso regression and random forest survival were used to screen variables related to the survival of advanced CRC patients received neoadjuvant therapy in the training group. Area under curves were adopted to evaluate the 1,3,5-year prediction value of the optimal model in three cohorts. Calibration curves were drawn to observe the prediction accuracy of the nomogram model. Decision curve analysis was used to assess the potential clinical value of the nomogram model.ResultsA total of 1833 subjects were enrolled in this study. After random allocation, 1055 cases of the SEER database served as the training group, 704 cases as the testing group and 74 patients from our center as the external validation group. Variables were screened by univariate cox regression used to construct a nomogram survival prediction model, including M, age, chemotherapy, CEA, perineural invasion, tumor size, LODDS, liver metastasis and radiation. The AUCs of the model for predicting 1-year OS in the training group, testing and validation group were 0.765 (0.703,0.827), 0.772 (0.697,0.847) and 0.742 (0.601,0.883), predicting 3-year OS were 0.761 (0.725,0.780), 0.742 (0.699,0.785), 0.733 (0.560,0.905) and 5-year OS were 0.742 (0.711,0.773), 0.746 (0.709,0.783), 0.838 (0.670,0.980), respectively. The calibration curves showed the difference between prediction probability of the model and the actual survival was not significant in three cohorts and the decision curve analysis revealed the practice clinical application value. And the prediction value of model was better for young CRC than older CRC patients.ConclusionA nomogram model including LODDS for the prognosis of advanced CRC received neoadjuvant therapy was constructed and verified based on the SEER database and single center practice. The accuracy and potential clinical application value of the model performed well, and the model had better predictive value for EOCRC than LOCRC.

Cross-attention enables deep learning on limited omics-imaging-clinical data of 130 lung cancer patients

Deep-learning tools that extract prognostic factors derived from multi-omics data have recently contributed to individualized predictions of survival outcomes. However, the limited size of integrated omics-imaging-clinical datasets poses challenges. Here, we propose two biologically interpretable and robust deep-learning architectures for survival prediction of non-small cell lung cancer (NSCLC) patients, learning simultaneously from computed tomography (CT) scan images, gene expression data, and clinical information. The proposed models integrate patient-specific clinical, transcriptomic, and imaging data and incorporate Kyoto Encyclopedia of Genes and Genomes (KEGG) and Reactome pathway information, adding biological knowledge within the learning process to extract prognostic gene biomarkers and molecular pathways. While both models accurately stratify patients in high- and low-risk groups when trained on a dataset of only 130 patients, introducing a cross-attention mechanism in a sparse autoencoder significantly improves the performance, highlighting tumor regions and NSCLC-related genes as potential biomarkers and thus offering a significant methodological advancement when learning from small imaging-omics-clinical samples.

Multimodal Machine Learning for Prognosis and Survival Prediction in Renal Cell Carcinoma Patients: A Two-Stage Framework with Model Fusion and Interpretability Analysis

Current medical limitations in predicting cancer survival status and time necessitate advancements beyond traditional methods and physical indicators. This research introduces a novel two-stage prognostic framework for renal cell carcinoma, addressing the inadequacies of existing diagnostic approaches. In the first stage, the framework accurately predicts the survival status (alive or deceased) with metrics Accuracy, Precision, Recall, and F1 score to evaluate the effects of the classification results, while the second stage focuses on forecasting the future survival time of deceased patients with Root Mean Square Error and Mean Absolute Error to evaluate the regression results. Leveraging popular machine learning models, such as Adaptive Boosting, Extra Trees, Gradient Boosting, Random Forest, and Extreme Gradient Boosting, along with fusion models like Voting, Stacking, and Blending, our approach significantly improves prognostic accuracy as shown in our experiments. The novelty of our research lies in the integration of a logistic regression meta-model for interpreting the blending model’s predictions, enhancing transparency. By the SHapley Additive exPlanations’ interpretability, we provide insights into variable contributions, aiding understanding at both global and local levels. Through modal segmentation and multimodal fusion applied to raw data from the Surveillance, Epidemiology, and End Results program, we enhance the precision of renal cell carcinoma prognosis. Our proposed model provides an interpretable analysis of model predictions, highlighting key variables influencing classification and regression decisions in the two-stage renal cell carcinoma prognosis framework. By addressing the black-box problem inherent in machine learning, our proposed model helps healthcare practitioners with a more reliable and transparent basis for applying machine learning in cancer prognostication.

Triple-negative breast cancer survival prediction: population-based research using the SEER database and an external validation cohort.

Triple-negative breast cancer (TNBC) is linked to a poorer outlook, heightened aggressiveness relative to other breast cancer variants, and limited treatment choices. The absence of conventional treatment methods makes TNBC patients susceptible to metastasis. The objective of this research was to assess the clinical and pathological traits of TNBC patients, predict the influence of risk elements on their outlook, and create a prediction model to assist doctors in treating TNBC patients and enhancing their prognosis. We included 23,394 individuals with complete baseline clinical data and survival information who were diagnosed with primary TNBC between 2010 and 2015 based on the SEER database. External validation utilised a group from The Affiliated Lihuili Hospital of Ningbo University. Independent risk factors linked to TNBC prognosis were identified through univariate, multivariate, and least absolute shrinkage and selection operator regression methods. These characteristics were chosen as parameters to develop 3- and 5-year overall survival (OS) and breast cancer-specific survival (BCSS) nomogram models. Model accuracy was assessed using calibration curves, consistency indices (C-indices), receiver operating characteristic curves (ROCs), and decision curve analyses (DCAs). Finally, TNBC patients were divided into groups of high, medium, and low risk, employing the nomogram model for conducting a Kaplan-Meier survival analysis. In the training cohort, variables such as age at diagnosis, marital status, grade, T stage, N stage, M stage, surgery, radiation, and chemotherapy were linked to OS and BCSS. For the nomogram, the C-indices stood at 0.762, 0.747, and 0.764 in forecasting OS across the training, internal validation, and external validation groups, respectively. Additionally, the C-index values for the training, internal validation, and external validation groups in BCSS prediction stood at 0.793, 0.755, and 0.811, in that order. The findings revealed that the calibration of our nomogram model was successful, and the time-variant ROC curves highlighted its effectiveness in clinical settings. Ultimately, the clinical DCA showcased the prospective clinical advantages of the suggested model. Furthermore, the online version was simple to use, and nomogram classification may enhance the differentiation of TNBC prognosis and distinguish risk groups more accurately. These nomograms are precise tools for assessing risk in patients with TNBC and forecasting survival. They can help doctors identify prognostic markers and create more effective treatment plans for patients with TNBC, providing more accurate assessments of their 3- and 5-year OS and BCSS.

Improving Prediction of Survival and Progression in Metastatic Non-Small Cell Lung Cancer After Immunotherapy Through Machine Learning of Circulating Tumor DNA.

To use modern machine learning approaches to enhance and automate the feature extraction from the longitudinal circulating tumor DNA (ctDNA) data and to improve the prediction of survival and disease progression, risk stratification, and treatment strategies for patients with 1L non-small cell lung cancer (NSCLC). Using IMpower150 trial data on patients with untreated metastatic NSCLC treated with atezolizumab and chemotherapies, we developed a machine learning algorithm to extract predictive features from ctDNA kinetics, improving survival and progression prediction. We analyzed kinetic data from 17 ctDNA summary markers, including cell-free DNA concentration, allele frequency, tumor molecules in plasma, and mutation counts. Three hundred and ninety-eight patients with ctDNA data (206 in training and 192 in validation) were analyzed. Our models outperformed existing workflow using conventional temporal ctDNA features, raising overall survival (OS) concordance index to 0.72 and 0.71 from 0.67 and 0.63 for C3D1 and C4D1, respectively, and substantially improving progression-free survival (PFS) to approximately 0.65 from the previous 0.54-0.58, a 12%-20% increase. Additionally, they enhanced risk stratification for patients with NSCLC, achieving clear OS and PFS separation. Distinct patterns of ctDNA kinetic characteristics (eg, baseline ctDNA markers, depth of ctDNA responses, and timing of ctDNA clearance, etc) were revealed across the risk groups. Rapid and complete ctDNA clearance appears essential for long-term clinical benefit. Our machine learning approach offers a novel tool for analyzing ctDNA kinetics, extracting critical features from longitudinal data, improving our understanding of the link between ctDNA kinetics and progression/mortality risks, and optimizing personalized immunotherapies for 1L NSCLC.

A decentralized federated learning-based cancer survival prediction method with privacy protection

BackgroundSurvival prediction is one of the crucial goals in precision medicine, as accurate survival assessment can aid physicians in selecting appropriate treatment for individual patients. To achieve this aim, extensive data must be utilized to train the prediction model and prevent overfitting. However, the collection of patient data for disease prediction is challenging due to potential variations in data sources across institutions and concerns regarding privacy and ownership issues in data sharing. To facilitate the integration of cancer data from different institutions without violating privacy laws, we developed a federated learning-based data integration framework called AdFed, which can be used to evaluate patients’ survival while considering the privacy protection problem by utilizing the decentralized federated learning technology and regularization method. ResultsAdFed was tested on different cancer datasets that contain the patients’ information from different institutions. The experimental results show that AdFed using distributed data can achieve better performance in cancer survival prediction (AUC = 0.605) than the compared federated-learning-based methods (average AUC = 0.554). Additionally, to assess the biological interpretability of our method, in the case study we list 10 identified genes related to liver cancer selected by AdFed, among which 5 genes have been proved by literature review. ConclusionsThe results indicate that AdFed outperforms better than other federated-learning-based methods, and the interpretable algorithm can select biologically significant genes and pathways while ensuring the confidentiality and integrity of data.

Smart Biosensor for Breast Cancer Survival Prediction Based on Multi-View Multi-Way Graph Learning.

Biosensors play a crucial role in detecting cancer signals by orchestrating a series of intricate biological and physical transduction processes. Among various cancers, breast cancer stands out due to its genetic underpinnings, which trigger uncontrolled cell proliferation, predominantly impacting women, and resulting in significant mortality rates. The utilization of biosensors in predicting survival time becomes paramount in formulating an optimal treatment strategy. However, conventional biosensors employing traditional machine learning methods encounter challenges in preprocessing features for the learning task. Despite the potential of deep learning techniques to automatically extract useful features, they often struggle to effectively leverage the intricate relationships between features and instances. To address this challenge, our study proposes a novel smart biosensor architecture that integrates a multi-view multi-way graph learning (MVMWGL) approach for predicting breast cancer survival time. This innovative approach enables the assimilation of insights from gene interactions and biosensor similarities. By leveraging real-world data, we conducted comprehensive evaluations, and our experimental results unequivocally demonstrate the superiority of the MVMWGL approach over existing methods.

A novel non-negative Bayesian stacking modeling method for Cancer survival prediction using high-dimensional omics data

BackgroundSurvival prediction using high-dimensional molecular data is a hot topic in the field of genomics and precision medicine, especially for cancer studies. Considering that carcinogenesis has a pathway-based pathogenesis, developing models using such group structures is a closer mimic of disease progression and prognosis. Many approaches can be used to integrate group information; however, most of them are single-model methods, which may account for unstable prediction.MethodsWe introduced a novel survival stacking method that modeled using group structure information to improve the robustness of cancer survival prediction in the context of high-dimensional omics data. With a super learner, survival stacking combines the prediction from multiple sub-models that are independently trained using the features in pre-grouped biological pathways. In addition to a non-negative linear combination of sub-models, we extended the super learner to non-negative Bayesian hierarchical generalized linear model and artificial neural network. We compared the proposed modeling strategy with the widely used survival penalized method Lasso Cox and several group penalized methods, e.g., group Lasso Cox, via simulation study and real-world data application.ResultsThe proposed survival stacking method showed superior and robust performance in terms of discrimination compared with single-model methods in case of high-noise simulated data and real-world data. The non-negative Bayesian stacking method can identify important biological signal pathways and genes that are associated with the prognosis of cancer.ConclusionsThis study proposed a novel survival stacking strategy incorporating biological group information into the cancer prognosis models. Additionally, this study extended the super learner to non-negative Bayesian model and ANN, enriching the combination of sub-models. The proposed Bayesian stacking strategy exhibited favorable properties in the prediction and interpretation of complex survival data, which may aid in discovering cancer targets.

Pathformer: a biological pathway informed transformer for disease diagnosis and prognosis using multi-omics data.

Multi-omics data provide a comprehensive view of gene regulation at multiple levels, which is helpful in achieving accurate diagnosis of complex diseases like cancer. However, conventional integration methods rarely utilize prior biological knowledge and lack interpretability. To integrate various multi-omics data of tissue and liquid biopsies for disease diagnosis and prognosis, we developed a biological pathway informed Transformer, Pathformer. It embeds multi-omics input with a compacted multi-modal vector and a pathway-based sparse neural network. Pathformer also leverages criss-cross attention mechanism to capture the crosstalk between different pathways and modalities. We first benchmarked Pathformer with 18 comparable methods on multiple cancer datasets, where Pathformer outperformed all the other methods, with an average improvement of 6.3%-14.7% in F1 score for cancer survival prediction, 5.1%-12% for cancer stage prediction, and 8.1%-13.6% for cancer drug response prediction. Subsequently, for cancer prognosis prediction based on tissue multi-omics data, we used a case study to demonstrate the biological interpretability of Pathformer by identifying key pathways and their biological crosstalk. Then, for cancer early diagnosis based on liquid biopsy data, we used plasma and platelet datasets to demonstrate Pathformer's potential of clinical applications in cancer screening. Moreover, we revealed deregulation of interesting pathways (e.g. scavenger receptor pathway) and their crosstalk in cancer patients' blood, providing potential candidate targets for cancer microenvironment study. Pathformer is implemented and freely available at https://github.com/lulab/Pathformer.