Dataset Information

AUC-based biomarker ensemble with an application on gene scores predicting low bone mineral density.

ABSTRACT:

Motivation

The area under the receiver operating characteristic (ROC) curve (AUC), long regarded as a 'golden' measure for the predictiveness of a continuous score, has propelled the need to develop AUC-based predictors. However, the AUC-based ensemble methods are rather scant, largely due to the fact that the associated objective function is neither continuous nor concave. Indeed, there is no reliable numerical algorithm identifying optimal combination of a set of biomarkers to maximize the AUC, especially when the number of biomarkers is large.

Results

We have proposed a novel AUC-based statistical ensemble methods for combining multiple biomarkers to differentiate a binary response of interest. Specifically, we propose to replace the non-continuous and non-convex AUC objective function by a convex surrogate loss function, whose minimizer can be efficiently identified. With the established framework, the lasso and other regularization techniques enable feature selections. Extensive simulations have demonstrated the superiority of the new methods to the existing methods. The proposal has been applied to a gene expression dataset to construct gene expression scores to differentiate elderly women with low bone mineral density (BMD) and those with normal BMD. The AUCs of the resulting scores in the independent test dataset has been satisfactory.

Conclusion

Aiming for directly maximizing AUC, the proposed AUC-based ensemble method provides an efficient means of generating a stable combination of multiple biomarkers, which is especially useful under the high-dimensional settings.

Contact

lutian@stanford.edu.

Supplementary information

Supplementary data are available at Bioinformatics online.

SUBMITTER: Zhao XG

PROVIDER: S-EPMC3198577 | biostudies-literature | 2011 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

AUC-based biomarker ensemble with an application on gene scores predicting low bone mineral density.

Zhao X G XG Dai W W Li Y Y Tian L L

Bioinformatics (Oxford, England) 20110909 21

<h4>Motivation</h4>The area under the receiver operating characteristic (ROC) curve (AUC), long regarded as a 'golden' measure for the predictiveness of a continuous score, has propelled the need to develop AUC-based predictors. However, the AUC-based ensemble methods are rather scant, largely due to the fact that the associated objective function is neither continuous nor concave. Indeed, there is no reliable numerical algorithm identifying optimal combination of a set of biomarkers to maximize ...[more]

PMID: 21908541

Similar Datasets

Project description:BackgroundHuman immunodeficiency virus (HIV) infected individuals may have osteoporosis. We aimed to evaluate the bone mineral density (BMD) in naïve antiretroviral (ARV) treated HIV positive patients comparing native Italian group (ItG) to a Migrants group (MiG) upon arrival in Italy.MethodsWe conducted a cross-sectional study on 83 HIV patients less than 50 years old. We used the dual-energy X-ray absorptiometry (DXA) within six months from the HIV diagnosis. Participants were categorized as having low BMD if the femoral neck or total lumbar spine Z-score was- 2 or less.ResultsMiG showed low BMD more often than ItG (37.5% vs.13.6%), especially for the female gender (16.7% vs. 0.0%). A low CD4 rate (<200 cells/μl) was most often detected in MiG than ItG. In particular, we found most often male Italians with abnormal CD4 than male migrants (67.8% vs. 33.3%) and vice versa for females (30.5% vs. 66.7%). We found an abnormal bone mineral density at the lumbar site. Low BMD at the lumbar site was more frequently observed in female migrants than female Italians. Both male and female migrants had a Z-score value significantly lower than male and female Italians, respectively. By logistic regression low vitamin-D level was positively correlated to low BMD in ItG only. All data were verified and validated using a triple code identifier.ConclusionsBoth DXA and vitamin-D evaluation should be offered after the diagnosis of HIV infection. Lumbar site low BMD is an initial condition of bone loss in HIV young patients, especially in female migrants. Vitamin D levels and supplementation may be considered after HIV diagnosis independently of age to improve bone health.HighlightsThis study evaluates the frequency of bone mineral density in HIV positive patients naive to antiretroviral therapy. It compares the density of the native Italian population with that of HIV Migrants upon arrival in Italy. The results show that HIV positive migrants, even if younger than 50 years of age, are at risk for osteoporosis, especially if they are female.

Dataset Information

AUC-based biomarker ensemble with an application on gene scores predicting low bone mineral density.

Motivation

Results

Conclusion

Contact

Supplementary information

Publications

AUC-based biomarker ensemble with an application on gene scores predicting low bone mineral density.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets