Dataset Information

The application of naive Bayes model averaging to predict Alzheimer's disease from genome-wide data.

ABSTRACT: Predicting patient outcomes from genome-wide measurements holds significant promise for improving clinical care. The large number of measurements (eg, single nucleotide polymorphisms (SNPs)), however, makes this task computationally challenging. This paper evaluates the performance of an algorithm that predicts patient outcomes from genome-wide data by efficiently model averaging over an exponential number of naive Bayes (NB) models.This model-averaged naive Bayes (MANB) method was applied to predict late onset Alzheimer's disease in 1411 individuals who each had 312,318 SNP measurements available as genome-wide predictive features. Its performance was compared to that of a naive Bayes algorithm without feature selection (NB) and with feature selection (FSNB).Performance of each algorithm was measured in terms of area under the ROC curve (AUC), calibration, and run time.The training time of MANB (16.1 s) was fast like NB (15.6 s), while FSNB (1684.2 s) was considerably slower. Each of the three algorithms required less than 0.1 s to predict the outcome of a test case. MANB had an AUC of 0.72, which is significantly better than the AUC of 0.59 by NB (p<0.00001), but not significantly different from the AUC of 0.71 by FSNB. MANB was better calibrated than NB, and FSNB was even better in calibration. A limitation was that only one dataset and two comparison algorithms were included in this study.MANB performed comparatively well in predicting a clinical outcome from a high-dimensional genome-wide dataset. These results provide support for including MANB in the methods used to predict outcomes from large, genome-wide datasets.

SUBMITTER: Wei W

PROVIDER: S-EPMC3128400 | biostudies-literature | 2011 Jul-Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

The application of naive Bayes model averaging to predict Alzheimer's disease from genome-wide data.

Wei Wei W Visweswaran Shyam S Cooper Gregory F GF

Journal of the American Medical Informatics Association : JAMIA 20110701 4

<h4>Objective</h4>Predicting patient outcomes from genome-wide measurements holds significant promise for improving clinical care. The large number of measurements (eg, single nucleotide polymorphisms (SNPs)), however, makes this task computationally challenging. This paper evaluates the performance of an algorithm that predicts patient outcomes from genome-wide data by efficiently model averaging over an exponential number of naive Bayes (NB) models.<h4>Design</h4>This model-averaged naive Baye ...[more]

PMID: 21672907

Similar Datasets

Project description:Ranking and identifying biomarkers that are associated with disease from genome-wide measurements holds significant promise for understanding the genetic basis of common diseases. The large number of single nucleotide polymorphisms (SNPs) in genome-wide studies (GWAS), however, makes this task computationally challenging when the ranking is to be done in a multivariate fashion. This paper evaluates the performance of a multivariate graph-based method called label propagation (LP) that efficiently ranks SNPs in genome-wide data.The performance of LP was evaluated on a synthetic dataset and two late onset Alzheimer's disease (LOAD) genome-wide datasets, and the performance was compared to that of three control methods. The control methods included chi squared, which is a commonly used univariate method, as well as a Relief method called SWRF and a sparse logistic regression (SLR) method, which are both multivariate ranking methods. Performance was measured by evaluating the top-ranked SNPs in terms of classification performance, reproducibility between the two datasets, and prior evidence of being associated with LOAD.On the synthetic data LP performed comparably to the control methods. On GWAS data, LP performed significantly better than chi squared and SWRF in classification performance in the range from 10 to 1000 top-ranked SNPs for both datasets, and not significantly different from SLR. LP also had greater ranking reproducibility than chi squared, SWRF, and SLR. Among the 25 top-ranked SNPs that were identified by LP, there were 14 SNPs in one dataset that had evidence in the literature of being associated with LOAD, and 10 SNPs in the other, which was higher than for the other methods.LP performed considerably better in ranking SNPs in two high-dimensional genome-wide datasets when compared to three control methods. It had better performance in the evaluation measures we used, and is computationally efficient to be applied practically to data from genome-wide studies. These results provide support for including LP in the methods that are used to rank SNPs in genome-wide datasets.

Project description:Characterizing the distribution of effects from genome-wide genotyping data is crucial for understanding important aspects of the genetic architecture of complex traits, such as number or proportion of non-null loci, average proportion of phenotypic variance explained per non-null effect, power for discovery, and polygenic risk prediction. To this end, previous work has used effect-size models based on various distributions, including the normal and normal mixture distributions, among others. In this paper we propose a scale mixture of two normals model for effect size distributions of genome-wide association study (GWAS) test statistics. Test statistics corresponding to null associations are modeled as random draws from a normal distribution with zero mean; test statistics corresponding to non-null associations are also modeled as normal with zero mean, but with larger variance. The model is fit via minimizing discrepancies between the parametric mixture model and resampling-based nonparametric estimates of replication effect sizes and variances. We describe in detail the implications of this model for estimation of the non-null proportion, the probability of replication in de novo samples, the local false discovery rate, and power for discovery of a specified proportion of phenotypic variance explained from additive effects of loci surpassing a given significance threshold. We also examine the crucial issue of the impact of linkage disequilibrium (LD) on effect sizes and parameter estimates, both analytically and in simulations. We apply this approach to meta-analysis test statistics from two large GWAS, one for Crohn's disease (CD) and the other for schizophrenia (SZ). A scale mixture of two normals distribution provides an excellent fit to the SZ nonparametric replication effect size estimates. While capturing the general behavior of the data, this mixture model underestimates the tails of the CD effect size distribution. We discuss the implications of pervasive small but replicating effects in CD and SZ on genomic control and power. Finally, we conclude that, despite having very similar estimates of variance explained by genotyped SNPs, CD and SZ have a broadly dissimilar genetic architecture, due to differing mean effect size and proportion of non-null loci.

Dataset Information

The application of naive Bayes model averaging to predict Alzheimer's disease from genome-wide data.

Publications

The application of naive Bayes model averaging to predict Alzheimer's disease from genome-wide data.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets