nondetects

nondetects models qPCR non-detects—reactions that fail to produce a minimum amount of signal—as missing not at random (MNAR) to reduce bias in estimating absolute and differential gene expression levels.

Key Features:

• Modeling Non-Detects as Missing Data: Recognizes qPCR non-detects as missing not at random (MNAR) and models the missing data mechanism instead of treating them as missing completely at random.
• Bias Reduction: Directly modeling non-detects reduces bias in estimation of both absolute and differential gene expression levels.
• Implementation in R: Implemented as an R package for integration into qPCR data analysis workflows.

Scientific Applications:

• Differential Gene Expression Analysis: Improves identification of genes that are upregulated or downregulated across conditions by accounting for MNAR non-detects.
• Absolute Quantification Studies: Enhances estimation of target gene copy numbers in qPCR experiments by reducing bias from non-detects.

Methodology:

Develops and applies a statistical model that accounts for the MNAR nature of qPCR non-detects and incorporates this model into the data analysis pipeline.