Sensei

Sensei estimates sample sizes required to detect differences in cell-type abundances in single-cell studies, supporting statistical design of cancer evolution and metastasis research using single-cell RNA sequencing and mass cytometry.

Key Features:

• Sample Size Estimation: Estimates the number of samples and cells required to detect changes in cell-type abundances between two groups in single-cell studies.
• Statistical Modeling: Models cell abundances using a beta-binomial distribution and expands traditional t-tests to account for single-cell variability and heterogeneity.
• Single-Cell Technology Support: Leverages data from single-cell RNA sequencing and mass cytometry for abundance and expression analyses.
• Practical Guidelines and Evaluation: Provides guidelines based on evaluations across over 20 cell types in more than 30 cancer types using data from the Cancer Cell Atlas (TCGA) and prior single-cell studies.

Scientific Applications:

• Study Design for Abundance Changes: Designs statistically powered experiments to detect cell-type abundance differences in heterogeneous cancer samples.
• Cancer Evolution and Metastasis Research: Informs sample-size decisions for investigations of cancer evolution and metastasis at the cellular level.

Methodology:

Extends traditional statistical tests such as t-tests by modeling cell abundances with a beta-binomial distribution.