MOTIVATION: Quantitative real-time PCR (qPCR) is one of the most widely used methods to measure gene expression. Despite extensive research in qPCR laboratory protocols, normalization and statistical analysis, little attention has been given to qPCR non-detects-those reactions failing to produce a minimum amount of signal. RESULTS: We show that the common methods of handling qPCR non-detects lead to biased inference. Furthermore, we show that non-detects do not represent data missing completely at random and likely represent missing data occurring not at random. We propose a model of the missing data mechanism and develop a method to directly model non-detects as missing data. Finally, we show that our approach results in a sizeable reduction in bias when estimating both absolute and differential gene expression. AVAILABILITY AND IMPLEMENTATION: The proposed algorithm is implemented in the R package, nondetects. This package also contains the raw data for the three example datasets used in this manuscript. The package is freely available at http://mnmccall.com/software and as part of the Bioconductor project.

Affymetrix GeneChip microarrays are the most widely used high-throughput technology to measure gene expression, and a wide variety of preprocessing methods have been developed to transform probe intensities reported by a microarray scanner into gene expression estimates. There have been numerous comparisons of these preprocessing methods, focusing on the most common analyses-detection of differential expression and gene or sample clustering. Recently, more complex multivariate analyses, such as gene co-expression, differential co-expression, gene set analysis and network modeling, are becoming more common; however, the same preprocessing methods are typically applied. In this article, we examine the effect of preprocessing methods on some of these multivariate analyses and provide guidance to the user as to which methods are most appropriate.