Feature selection has become the essential step in biomarker discovery from high-dimensional genomics data. It is recognized that different feature selection techniques may result in different set of biomarkers, that is, different groups of genes highly correlated to a given pathological condition, but few direct comparisons exist which quantify these differences in a systematic way. In this paper, we propose a general methodology for comparing the outcomes of different selection techniques in the context of biomarker discovery. The comparison is carried out along two dimensions: (i) measuring the similarity/dissimilarity of selected gene sets; (ii) evaluating the implications of these differences in terms of both predictive performance and stability of selected gene sets. As a case study, we considered three benchmarks deriving from DNA microarray experiments and conducted a comparative analysis among eight selection methods, representatives of different classes of feature selection techniques. Our results show that the proposed approach can provide useful insight about the pattern of agreement of biomarker discovery techniques.
The massive production and spread of biomedical data around the web introduces new challenges related to identify computational approaches for providing quality search and browsing of web resources. This papers presents BioCloud Search EnGene (BSE), a cloud application that facilitates searching and integration of the many layers of biological information offered by public large-scale genomic repositories. Grounding on the concept of dataspace, BSE is built on top of a cloud platform that severely curtails issues associated with scalability and performance. Like popular online gene portals, BSE adopts a gene-centric approach: researchers can find their information of interest by means of a simple “Google-like” query interface that accepts standard gene identification as keywords. We present BSE architecture and functionality and discuss how our strategies contribute to successfully tackle big data problems in querying gene-based web resources. BSE is publically available at: http://biocloud-unica.appspot.com/
HER2+ breast cancer (BC) is an aggressive subtype representing a genetically and biologically heterogeneous group of tumors resulting in variable prognosis and treatment response to HER2-targeted therapies according to estrogen (ER) and progesterone receptor (PR) expression. The relationship with androgen receptors (AR), a member of the steroid hormone’s family, is unwell known in BC. The present study aims to evaluate the prognostic impact of AR expression in HER2+ BC subtypes. A total of 695 BCs were selected and reviewed, AR, ER, PR and HER2 expression in tumor cells were examined by immunohistochemical method, and the SISH method was used in case of HER2 with equivocal immunohistochemical score (2+). A high prevalence of AR expression (91.5%) in BC HER+ was observed, with minimal differences between luminal and non-luminal tumor. According to steroid receptor expression, tumors were classified in four subgroups, including BC luminal and non-luminal HER2+ expressing or not AR. The luminal BC HER2 + AR+ was associated with lower histological grade, lower tumor size, higher PR expression and lower HER2 intensity of expression (2+). Also, the non-luminal tumors AR+ showed lower tumor size and lower prognostic stage but frequently higher grade and higher HER2 intensity of expression (3+). These findings should suggest a different progression of luminal and non-luminal tumors, both expressing AR, and allow us to speculate that the molecular mechanisms of AR, involved in the biology of BC HER2 + AR+, differ in relation to ER and PR expression. Moreover, AR expression may be a useful predictor of prognosis for overall survival (OS) in HER2+ BC subtypes. Our findings suggest that AR expression evaluation in clinical practice could be utilized in clinical oncology to establish different aggressiveness in BC HER2+ subtypes.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.