Dataset Information

Iteratively refining breast cancer intrinsic subtypes in the METABRIC dataset.

ABSTRACT: BACKGROUND:Multi-gene lists and single sample predictor models have been currently used to reduce the multidimensional complexity of breast cancers, and to identify intrinsic subtypes. The perceived inability of some models to deal with the challenges of processing high-dimensional data, however, limits the accurate characterisation of these subtypes. Towards the development of robust strategies, we designed an iterative approach to consistently discriminate intrinsic subtypes and improve class prediction in the METABRIC dataset. FINDINGS:In this study, we employed the CM1 score to identify the most discriminative probes for each group, and an ensemble learning technique to assess the ability of these probes on assigning subtype labels using 24 different classifiers. Our analysis is comprised of an iterative computation of these methods and statistical measures performed on a set of over 2000 samples. The refined labels assigned using this iterative approach revealed to be more consistent and in better agreement with clinicopathological markers and patients' overall survival than those originally provided by the PAM50 method. CONCLUSIONS:The assignment of intrinsic subtypes has a significant impact in translational research for both understanding and managing breast cancer. The refined labelling, therefore, provides more accurate and reliable information by improving the source of fundamental science prior to clinical applications in medicine.

SUBMITTER: Milioli HH

PROVIDER: S-EPMC4712506 | biostudies-literature | 2016

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Iteratively refining breast cancer intrinsic subtypes in the METABRIC dataset.

Milioli Heloisa H HH Vimieiro Renato R Tishchenko Inna I Riveros Carlos C Berretta Regina R Moscato Pablo P

BioData mining 20160113

<h4>Background</h4>Multi-gene lists and single sample predictor models have been currently used to reduce the multidimensional complexity of breast cancers, and to identify intrinsic subtypes. The perceived inability of some models to deal with the challenges of processing high-dimensional data, however, limits the accurate characterisation of these subtypes. Towards the development of robust strategies, we designed an iterative approach to consistently discriminate intrinsic subtypes and improv ...[more]

PMID: 26770261

Similar Datasets

Project description:BackgroundThe prediction of breast cancer intrinsic subtypes has been introduced as a valuable strategy to determine patient diagnosis and prognosis, and therapy response. The PAM50 method, based on the expression levels of 50 genes, uses a single sample predictor model to assign subtype labels to samples. Intrinsic errors reported within this assay demonstrate the challenge of identifying and understanding the breast cancer groups. In this study, we aim to: a) identify novel biomarkers for subtype individuation by exploring the competence of a newly proposed method named CM1 score, and b) apply an ensemble learning, as opposed to the use of a single classifier, for sample subtype assignment. The overarching objective is to improve class prediction.Methods and findingsThe microarray transcriptome data sets used in this study are: the METABRIC breast cancer data recorded for over 2000 patients, and the public integrated source from ROCK database with 1570 samples. We first computed the CM1 score to identify the probes with highly discriminative patterns of expression across samples of each intrinsic subtype. We further assessed the ability of 42 selected probes on assigning correct subtype labels using 24 different classifiers from the Weka software suite. For comparison, the same method was applied on the list of 50 genes from the PAM50 method.ConclusionsThe CM1 score portrayed 30 novel biomarkers for predicting breast cancer subtypes, with the confirmation of the role of 12 well-established genes. Intrinsic subtypes assigned using the CM1 list and the ensemble of classifiers are more consistent and homogeneous than the original PAM50 labels. The new subtypes show accurate distributions of current clinical markers ER, PR and HER2, and survival curves in the METABRIC and ROCK data sets. Remarkably, the paradoxical attribution of the original labels reinforces the limitations of employing a single sample classifiers to predict breast cancer intrinsic subtypes.

Project description:BackgroundThe discovery of breast cancer subtypes and subsequent development of treatments aimed at them has allowed for a great reduction in the mortality of breast cancer. But despite this progress, tumors with similar characteristics that belong to the same subtype continue to respond differently to the same treatment. Five subtypes of breast cancer, namely intrinsic subtypes, have been characterized to date based on their gene expression profiles. Among other characteristics, subtypes vary in their degree of intra-subtype heterogeneity. It is not clear, however, whether this heterogeneity is shared across all tumor traits. It is also unclear whether individual traits can be highly heterogeneous among a majority of homogeneous traits.ResultsWe employ network theory to uncover gene modules and accordingly consider them as tumor traits, which capture shared biological processes among the subtypes. We then use the ?-diversity metric from ecology to quantify the heterogeneity in these gene modules. In doing so, we show that breast cancer heterogeneity is contained in gene modules and that this modular heterogeneity increases monotonically across the subtypes. We identify a core of two modules that are shared among all subtypes which contain nucleosome assembly and mammary morphogenesis genes, and a number of modules that are specific to subtypes. This modular heterogeneity, which increases with global heterogeneity, relates to tumor aggressiveness. Indeed, we observe that Luminal A, the most treatable of subtypes, has the lowest modular heterogeneity whereas the Basal-like subtype, which is among the hardest to treat, has the highest. Furthermore, our analysis shows that a higher degree of global heterogeneity does not imply higher heterogeneity for all modules, as Luminal B shows the highest heterogeneity for core modules.ConclusionsOverall, modular heterogeneity provides a framework with which to dissect cancer heterogeneity and better understand its underpinnings, thereby ultimately advancing our knowledge towards a more effective personalized cancer therapy.

Dataset Information

Iteratively refining breast cancer intrinsic subtypes in the METABRIC dataset.

Publications

Iteratively refining breast cancer intrinsic subtypes in the METABRIC dataset.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure