Marker Enrichment Modeling (MEM)

Marker Enrichment Modeling (MEM) software is a quantitative system for objectively labeling and characterizing complex cell populations by quantifying feature enrichment and reporting a text label. MEM label reporting provides a new way to objectively and quantitatively communicate cell type in studies of complex, dynamic tissue microenvironments, such as stem cell niches, immune responses, and human tumors. Since MEM labels are quantitative, human and machine readable, they provide a key missing tool needed for machine learning of cell identity in applications such as cell biology, pathology, assessment of quality over time for cell based products, and biomarker discovery.  

Core of this software is the MEM equation and R package that calculates and visualizes cell population-specific feature enrichment. The equation combines magnitude and variance information in order to determine which measured features are enriched on, or specific to, each input population compared to a reference population. The R package takes data files as input that contain the single-cell feature expression information and the cluster designation for each cell. It outputs MEM scores that quantify the positive or negative enrichment of each feature on each population and also visualizes the results in a heatmap.

MEM Software is provided free of charge to academic and not-for-profit institutions via our academic online licensing process for non-commercial research purposes. Once you have completed checkout with this item, you will receive an e-mail with download instructions.