DepecheR

DepecheR identifies traits in high-dimensional datasets that separate groups through sparse K-means clustering and subsequent prediction of group outcomes based on cluster distributions.

Key Features:

• Sparse K-means Clustering: Implements sparse K-means to perform clustering that identifies groups defined by a sparse number of variables.
• Group Outcome Prediction: Uses generated cluster distributions to predict outcomes for groups of individuals.
• Data Complexity Reduction: Identifies clusters with separating information to emphasize only the data that matters by reducing dimensionality.

Scientific Applications:

• High-Dimensional Data Analysis: Identifies essential phenotypic elements that separate groups in high-dimensional entities.
• Trait Identification: Discovers separating traits in datasets to distinguish between biological or clinical groups.

Methodology:

DepecheR performs clustering using an implementation of sparse K-means, then uses the generated clusters to predict outcomes of groups of individuals based on their distribution of observations in the different clusters.