BackgroundMicroarray data have been used for gene signature selection to predict clinical outcomes. Many studies have attempted to identify factors that affect models' performance with only little success. Fine-tuning of model parameters and optimizing each step of the modeling process often results in over-fitting problems without improving performance.ResultsWe propose a quantitative measurement, termed consistency degree, to detect the correlation between disease endpoint and gene expression profile. Different endpoints were shown to have different consistency degrees to gene expression profiles. The validity of this measurement to estimate the consistency was tested with significance at a p-value less than 2.2e-16 for all of the studied endpoints. According to the consistency degree score, overall survival milestone outcome of multiple myeloma was proposed to extend from 730 days to 1561 days, which is more consistent with gene expression profile.ConclusionFor various clinical endpoints, the maximum predictive powers of different microarray-based models are limited by the correlation between endpoint and gene expression profile of disease samples as indicated by the consistency degree score. In addition, previous defined clinical outcomes can also be reassessed and refined more coherent according to related disease gene expression profile. Our findings point to an entirely new direction for assessing the microarray-based predictive models and provide important information to gene signature based clinical applications.

This is an editorial report of the supplement to BMC Genomics that includes 15 papers selected from the BIOCOMP'10 - The 2010 International Conference on Bioinformatics & Computational Biology as well as other sources with a focus on genomics studies.BIOCOMP'10 was held on July 12-15 in Las Vegas, Nevada. The congress covered a large variety of research areas, and genomics was one of the major focuses because of the fast development in this field. We set out to launch a supplement to BMC Genomics with manuscripts selected from this congress and invited submissions. With a rigorous peer review process, we selected 15 manuscripts that showed work in cutting-edge genomics fields and proposed innovative methodology. We hope this supplement presents the current computational and statistical challenges faced in genomics studies, and shows the enormous promises and opportunities in the genomic future.

BackgroundMicroarray data have a high dimension of variables and a small sample size. In microarray data analyses, two important issues are how to choose genes, which provide reliable and good prediction for disease status, and how to determine the final gene set that is best for classification. Associations among genetic markers mean one can exploit information redundancy to potentially reduce classification cost in terms of time and money.ResultsTo deal with redundant information and improve classification, we propose a gene selection method, Recursive Feature Addition, which combines supervised learning and statistical similarity measures. To determine the final optimal gene set for prediction and classification, we propose an algorithm, Lagging Prediction Peephole Optimization. By using six benchmark microarray gene expression data sets, we compared Recursive Feature Addition with recently developed gene selection methods: Support Vector Machine Recursive Feature Elimination, Leave-One-Out Calculation Sequential Forward Selection and several others.ConclusionsOn average, with the use of popular learning machines including Nearest Mean Scaled Classifier, Support Vector Machine, Naive Bayes Classifier and Random Forest, Recursive Feature Addition outperformed other methods. Our studies also showed that Lagging Prediction Peephole Optimization is superior to random strategy; Recursive Feature Addition with Lagging Prediction Peephole Optimization obtained better testing accuracies than the gene selection method varSelRF.

BackgroundIn microarray experiments with small sample sizes, it is a challenge to estimate p-values accurately and decide cutoff p-values for gene selection appropriately. Although permutation-based methods have proved to have greater sensitivity and specificity than the regular t-test, their p-values are highly discrete due to the limited number of permutations available in very small sample sizes. Furthermore, estimated permutation-based p-values for true nulls are highly correlated and not uniformly distributed between zero and one, making it difficult to use current false discovery rate (FDR)-controlling methods.ResultsWe propose a model-based information sharing method (MBIS) that, after an appropriate data transformation, utilizes information shared among genes. We use a normal distribution to model the mean differences of true nulls across two experimental conditions. The parameters of the model are then estimated using all data in hand. Based on this model, p-values, which are uniformly distributed from true nulls, are calculated. Then, since FDR-controlling methods are generally not well suited to microarray data with very small sample sizes, we select genes for a given cutoff p-value and then estimate the false discovery rate.ConclusionSimulation studies and analysis using real microarray data show that the proposed method, MBIS, is more powerful and reliable than current methods. It has wide application to a variety of situations.