CombiROC takes advantage of the combinatorial analysis and ROC curves although these methods are being used in medicine and other areas for many decades, they generally lack an easy-to-use interface that researchers without programming skills could use to analyze data and to make plots. In this study we describe CombiROC, an easy-to-use tool implemented as a web application to accurately determine optimal combinations of markers from diverse complex omics data. Moreover, confirmation analyses are necessary to evaluate the performance of candidate biomarker panels in order to avoid the risk of over-fitting, which request to conduct a validation step on an independent cohort, or alternatively, by means of cross validation or bootstrapping 18, 19. Even if quite a number of statistical methods for combining biomarkers exists, their application in clinical environment is still a prerogative of analytically skilled researchers mainly due to the difficulty to extrapolate simplified, standardized and interpretable result from these complex statistical strategies 15, 16, 17. Literature provides various statistical modeling strategies to combine biomarkers among them, threshold-based 9, 10, logistic regression 11, 12 and tree-based 13 methods are probably the most utilized whereas techniques such as Support Vector Machines 14 represent helpful tools in many high-dimensional problems. Moreover, it is now generally accepted that single markers do not achieve sufficient sensitivity and specificity for translation into diagnostic setting, thereby clinical decision-making process benefits from combining signatures to improve clinical performances: this trend clearly emerged over the past few years 4, 5, 6, 7, 8. Unfortunately, most of them could be too expensive to achieve the specificity and/or sensitivity needed for diagnostics and routine clinical practices. In the last years the identification of disease-associated signatures has been accelerated by the use of high-throughput omics techniques 1, 2, 3. Accurate markers are powerful tools to help clinicians in the choice of the most appropriate treatment, ultimately improving a personalized patient management. Moreover, it can be used to foresee the more likely outcome of the disease, monitor its progression and predict the response to a given therapy. ![]() In the area of diagnostic medicine, a biomarker is often used as a tool to identify subjects with a disease, or at high risk of developing the disease. Powerful biomarkers have become and will continue to be important tools in diagnostic, clinical and research settings. CombiROC is a novel tool for the scientific community freely available at. The application was validated with published data, confirming the marker combination already originally described or even finding new ones. ![]() CombiROC was designed without hard-coded thresholds, allowing a custom fit to each specific data: this dramatically reduces the computational burden and lowers the false negative rates given by fixed thresholds. Leaving to the user full control on initial selection stringency, CombiROC computes sensitivity and specificity for all markers combinations, performances of best combinations and ROC curves for automatic comparisons, all visualized in a graphic interface. With CombiROC data from different domains, such as proteomics and transcriptomics, can be analyzed using sensitivity/specificity filters: the number of candidate marker panels rising from combinatorial analysis is easily optimized bypassing limitations imposed by the nature of different experimental approaches. We developed a user-friendly tool, called CombiROC, to help researchers accurately determine optimal markers combinations from diverse omics methods. ![]() ![]() The selection of multimarker signatures is a complicated process that requires integration of data signatures with sophisticated statistical methods. Diagnostic accuracy can be improved considerably by combining multiple markers, whose performance in identifying diseased subjects is usually assessed via receiver operating characteristic (ROC) curves.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |