SLDAssay: A software package and web tool for analyzing limiting dilution assays.

Area of Science:

• Infectious disease research
• Computational biology
• Statistical modeling

Background:

• Serial limiting dilution (SLD) assays are crucial for quantifying target cells in infectious disease research.
• Existing software lacks advanced statistical methods for SLD assay data analysis.

Purpose of the Study:

• To introduce SLDAssay, a novel R software package and web tool for analyzing SLD assay data.
• To provide accurate estimation of target cell concentrations using maximum likelihood estimation (MLE) and bias-corrected (BC) MLE.
• To implement exact and asymptotic statistical methods for confidence intervals and goodness-of-fit testing.

Main Methods:

• Development of an R software package named SLDAssay.
• Implementation of maximum likelihood estimation (MLE) and bias-corrected (BC) MLE.
• Inclusion of exact and asymptotic methods for confidence intervals and goodness-of-fit p-values.
• Validation using existing data (Myers et al., 1994) and simulation studies.

Main Results:

• SLDAssay provides exact and asymptotic confidence intervals and goodness-of-fit p-values.
• Simulations show the BC MLE is less biased than the standard MLE.
• Exact methods demonstrate superior confidence interval coverage and lower type I error rates compared to asymptotic methods.

Conclusions:

• SLDAssay offers a comprehensive and accurate tool for SLD assay data analysis.
• The software's implementation of BC MLE and exact methods provides significant advantages over existing tools.
• SLDAssay enhances the reliability and precision of infectious disease research utilizing SLD assays.