PHENOSTAMP

PHENOSTAMP maps epithelial-mesenchymal transition (EMT) and mesenchymal-epithelial transition (MET) states in lung cancer by constructing an EMT-MET reference map and projecting single-cell mass cytometry data to characterize phenotypic transitions relevant to cancer progression and drug resistance.

Key Features:

• CCAST algorithm: Incorporates the Clustering Classification and Sorting Tree (CCAST) algorithm to classify and sort single-cell data for EMT and MET state identification.
• Neural network 2D visualization: Employs a neural network to generate optimal 2D projections that map single-cell data onto a reference EMT-MET state map.
• Mass cytometry time-course analysis: Leverages mass cytometry time-course data, including TGFβ-induced EMT and TGFβ withdrawal–induced MET, to resolve temporal state changes.
• Trajectory reconstruction (TRACER): Uses TRACER to reconstruct trajectories between cell states and compare EMT versus MET pathways.
• EMT-MET reference mapping: Constructs the EMT-MET PHENOtypic STAte MaP (PHENOSTAMP) from in vitro analyses as a benchmark for phenotypic profiling and projecting clinical samples.

Scientific Applications:

• Cancer progression analysis: Maps EMT and MET states to elucidate mechanisms underlying tumor progression and metastasis.
• Drug resistance studies: Characterizes cellular transitions at single-cell resolution to identify markers and states associated with drug resistance.
• Clinical sample characterization: Projects clinical samples onto the EMT-MET PHENOSTAMP reference map to assess and contextualize phenotypic profiles from patient specimens.

Methodology:

Applies CCAST for clustering/classification, a neural network for 2D projection, TRACER for trajectory reconstruction, and constructs an EMT-MET reference map from mass cytometry time-course data (TGFβ induction and withdrawal) for projecting clinical samples.