Statistical estimation & inference of cell counts from ELISPOT limiting dilution assays

Area of Science:

• Immunological assays and statistical analysis
• Quantitative immunology
• Biostatistics

Background:

• The ELISPOT assay is a key tool for enumerating immune cells, particularly antigen-specific memory cells.
• Accurate cell concentration estimation from ELISPOT spot counts is crucial for immunological studies.
• Automated methods are required to streamline the analysis of ELISPOT data.

Purpose of the Study:

• To develop and evaluate statistical methods for estimating cell concentrations from ELISPOT assay spot counts.
• To investigate the distribution of derived endpoints (antigen-specific memory cell percentages) for hypothesis testing.
• To provide a basis for comparing immune cell responses between groups using statistical tests.

Main Methods:

• Assumed three major distributions for observational cell counts in ELISPOT assays.
• Applied individual least squares (LS)/maximum likelihood and robust least squares (RLS) for parameter estimation.
• Investigated the distribution of study endpoints (percentage of antigen-specific memory cells per total IgG) and compared statistical tests (t-test, Wilcoxon Rank Sum test) via simulations and real data.

Main Results:

• Demonstrated that endpoint estimates across subjects are approximately identically distributed within groups under weak conditions.
• Confirmed the suitability of the t-test or Wilcoxon Rank Sum test for detecting group differences in ELISPOT assay data.
• Validated the proposed statistical methods through simulations and application to real experimental data.

Conclusions:

• Statistical methods, including LS, RLS, and standard hypothesis tests, can reliably analyze ELISPOT assay data.
• The distribution of derived endpoints allows for robust group comparisons in immunological studies.
• The findings support the use of automated analysis and statistical testing for ELISPOT assay data interpretation.