CamurWeb: a classification software and a large knowledge base for gene expression data of cancer

Abstract

BackgroundThe high growth of Next Generation Sequencing data currently demands new knowledge extraction methods. In particular, the RNA sequencing gene expression experimental technique stands out for case-control studies on cancer, which can be addressed with supervised machine learning techniques able to extract human interpretable models composed of genes, and their relation to the investigated disease. State of the art rule-based classifiers are designed to extract a single classification model, possibly composed of few relevant genes. Conversely, we aim to create a large knowledge base composed of many rule-based models, and thus determine which genes could be potentially involved in the analyzed tumor. This comprehensive and open access knowledge base is required to disseminate novel insights about cancer.ResultsWe propose CamurWeb, a new method and web-based software that is able to extract multiple and equivalent classification models in form of logic formulas (“if then” rules) and to create a knowledge base of these rules that can be queried and analyzed. The method is based on an iterative classification procedure and an adaptive feature elimination technique that enables the computation of many rule-based models related to the cancer under study. Additionally, CamurWeb includes a user friendly interface for running the software, querying the results, and managing the performed experiments. The user can create her profile, upload her gene expression data, run the classification analyses, and interpret the results with predefined queries. In order to validate the software we apply it to all public available RNA sequencing datasets from The Cancer Genome Atlas database obtaining a large open access knowledge base about cancer. CamurWeb is available at http://bioinformatics.iasi.cnr.it/camurweb.ConclusionsThe experiments prove the validity of CamurWeb, obtaining many classification models and thus several genes that are associated to 21 different cancer types. Finally, the comprehensive knowledge base about cancer and the software tool are released online; interested researchers have free access to them for further studies and to design biological experiments in cancer research.

GABRDpromotes the progression of breast cancer through CDK1-dependent cell cycle regulation

“…Recently, some studies have reported the GABRD gene expression levels are involved in the development of several malignancies. In particular, the mutation and overexpression of GABRD can cause the proliferation of tumor cells, especially in kidney renal clear cell carcinoma, thyroid carcinoma, and breast invasive carcinoma (9). Andrew M. Gross reported that increased GABRD levels that could lead to functional changes in the GABA A receptor might play a role in tumor cell differentiation (10).…”

Knowledge Generation with Rule Induction in Cancer Omics

“…Examples of successful usage of rule learning methods in cancer research have been reported both on small scale studies [37][38][39][40][41], as well as on large scale projects like TCGA [42][43][44].…”

“…In this latter case, specific interfaces, mediating the interaction between the user and the rule base need to be built. In two recent works, by applying rule induction to TCGA gene expression data [42,43], 21 rule sets, for as many different cancer datasets, were generated. In order to facilitate knowledge extraction from this sets, the authors proposed five different kinds of query: i) A gene list, to query the list of features used in the rule set of a given cancer type, ii) literals and conjunctions list, to retrieve the most relevant literals (e.g., geneA > 20) and conjunctions (e.g., geneA > 0.4 & geneB < 50) for a given cancer type, along with the number of instances classified by rules containing them, iii) rules list, to recover the list of all the rules derived for a given cancer type along with the corresponding measures of reliability, iv) literals statistic, to retrieve literals (i.e., genes) with occurrence in a given cancer within a specific frequency range, and v) gene pairs, to retrieve the gene pairs that occur more frequently in a rule set of a given cancer type.…”

LICIC: Less Important Components for Imbalanced Multiclass Classification

“…In the literature, works [18][19][20][21][22][23][24] have shown that, despite the high number of dimensions in RNA sequencing, DNA and Gene Expression Profiles datasets, an accurate classification is feasible if class imbalance is treated appropriately.…”