The above JavaScript is a basic function. If your browser does not support JavaScript, if the webpage function is not working properly, please enable the JavaScript status of the browser. Go to the content anchor

LOGO

How Long Can Cancer Patients Live?

:::HOME / LIFE SCIENCES / How Long Can Cancer Patients Live?
How Long Can Cancer Patients Live?

How long can cancer patients live? The prospect of survival for cancer patients depends on some key factors, such as the age at diagnosis, cancer type, cancer stage, cancer metastasis, and genetic features. As a result, the signature identification of biomarkers, such as genes and non-coding RNAs (ncRNAs), associated with survival time is urgently required for precision medicine. A biomarker signature is defined as a minimal subset of biomarkers that are maximally predictive with respect to survival time when they work cooperatively. Biostatistics approaches are commonly used for discovery of individual biomarkers, while machine learning approaches are appropriate for identifying biomarker signatures and establishing prediction models from a large dataset, e.g., expression profiles of genes/ncRNAs. However, the training expression profiles of cancer patients are usually characterized by high-dimensional samples and small datasets, which result in both curse of dimensionality and underdetermined problems.

We propose a novel evolutionary learning platform on complex biomedical data for identification of biomarker signatures while considering the underdetermined problem. For example, the proposed method identified 18 out of 332 miRNAs and achieved a correlation coefficient of 0.88 ± 0.01 and mean absolute error of 0.56 ± 0.03 years between real and estimated survival time of lung adenocarcinoma patients. The platform has been validated using miRNA/lncRNA expression profiles of patients with lung adenocarcinoma, glioblastoma multiforme, breast cancer, ovarian cancer and neuroblastoma from The Cancer Genome Atlas and Gene Expression Omnibus databases. The analysis of all signatures of genes/ncRNAs for cancer patients can help to develop drugs and gene targeted therapy.

 


How long can cancer patients live? The prospect of survival for cancer patients depends on some key factors, such as the age at diagnosis, cancer type, cancer stage, cancer metastasis, and genetic features. Precision medicine plays multiple roles in improving the long-term survival of patients. As a result, the identification of biomarkers associated with survival time is urgently required for precision medicine. Recent advances in next-generation sequencing and microarray technologies have resulted in great interest in non-coding RNAs (ncRNAs), including small non-coding RNAs, such as miRNAs and piRNAs, and long non-coding RNAs (lncRNAs), which have shown their significant roles in various cancers. In particular, the role of ncRNAs in evolution and genome functions is a new topic of interest in cancer research. A biomarker signature is defined as a minimal subset of biomarkers (e.g., ncRNAs) that are maximally predictive with respect to survival time when they work cooperatively. Generally, machine learning methods emphasize predictive results and statistical methods care about causal reasoning. Biostatistics approaches are commonly used for discovery of individual biomarkers, while machine learning approaches are appropriate for identifying biomarker signatures and establishing prediction models from a large dataset (e.g., expression profiles of ncRNAs). 
 
We therefore need to develop a machine learning technique in which the survival time can be predicted through the patient’s genetic, environmental, and lifestyle factors for personalized medicine. The identification of a m-biomarker signature from a large set of n candidate biomarkers is a combinatorial optimization problem of C(n, m), which relies on a specialized feature selection algorithm in cooperation with mathematical modeling. However, the training expression profiles of cancer patients are usually characterized by high-dimensional samples and small datasets, which result in both curse of dimensionality and underdetermined problems. The signature identification is a complex challenge that involves interdisciplinary fields, such as bioinformatics and artificial intelligence (AI). Despite the difficult but worthwhile challenge, we aim to develop machine learning methods for identifying signatures associated with the survival time of patients with cancers for AI in precision medicine.
 
We propose a novel evolutionary learning platform based on our intelligent evolutionary algorithm (IEA) for solving the intractable combinatorial optimization problem while considering the underdetermined problem. An optimal feature selection algorithm, inheritable bi-objective combinatorial genetic algorithm, which cooperates with support vector machines, identifies a small set of biomarkers while maximizing prediction accuracy. For example, the proposed method identified 18 out of 332 miRNAs using 10-fold cross-validation and achieved a correlation coefficient of 0.88 ± 0.01 and mean absolute error of 0.56 ± 0.03 years between real and estimated survival time of lung adenocarcinoma patients. Gene ontology annotation and pathway analysis of the 18-miRNA signature revealed its biological significance in cancer and cellular pathways. The signature identification platform has been validated using miRNA/lncRNA expression profiles of patients with lung adenocarcinoma, glioblastoma multiforme, breast cancer, ovarian cancer and neuroblastoma from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases.

The evolutionary learning platform can identify various biomarker signatures of genes/ncRNAs associated with survival time, cancer stage and cancer metastasis from the datasets of expression profiles in patients with cancers. Using the feedback mechanisms of the growing datasets (e.g., TCGA and GEO), the AI-based platform can incrementally refine the identified signatures and the derived prediction models. The identified signature for designing prediction models can comprise not only genes/ncRNAs but also biochemistry values and lifestyle factors for precision medicine and healthcare. The biomarker signatures could aid in the development of novel therapeutic approaches to the treatment of various cancers. Furthermore, the combined analysis of all signatures of genes/ncRNAs with their related pathways for cancer patients can help to develop drugs and gene targeted therapy.
 
 
Evolutionary learning platform for biomedical data.
Figure 1: Evolutionary learning platform for biomedical data.
 
The evolutionary learning platform for identifying biomarker signatures and designing prediction models for aiding decision making is extending to various types of biomedical datasets, such as gene/ncRNA expression profile, biomedical image, microbiota, electroencephalography, LC-MS/MS spectrum, public health database, etc. (Figure1). Differing from biostatistics and deep learning approaches, the evolutionary learning platform can easily incorporate biomedical domain knowledge into feature extraction and identify a small set of informative biomarkers. The achievements will be chosen to be presented during the COMPUTEX TAIPEI held on 28th May to 1st June, 2019 that attracts over 1,600 exhibitors and 40,000 international professional visitors attending every June in Taipei, Taiwan. The evolutionary learning platform-based computer aided diagnosis system for biomedical images won an Entrepreneurial Potential Award in the FITI (From Invention to Innovation) Program Competition of the Ministry of Science and Technology in 2017 (Figure2). The company (Doctor How Inc.), a derivative startup of National Chiao Tung University (NCTU), has the technical authorization of the evolutionary learning technique from NCTU for providing research service of AI medicine.
 
The computer aided diagnosis system won an Entrepreneurial Potential Award in the FITI Program Competition of the Ministry of Science and Technology.
Figure 2: The computer aided diagnosis system won an Entrepreneurial Potential Award in the FITI Program Competition of the Ministry of Science and Technology.

 

RELATED

STAY CONNECTED. SUBSCRIBE TO OUR NEWSLETTER.

Add your information below to receive daily updates.