rSuStaIn

The rSuStaIn package provides an R wrapper for the pySuStaIn Python package, which implements the Subtype and Stage Inference (SuStaIn) algorithm. SuStaIn is an unsupervised learning algorithm that identifies disease subtypes and stages from cross-sectional data.

Features

This package provides R wrappers for the three main SuStaIn algorithms:

• Ordinal SuStaIn (run_OSA): For ordinal/scored biomarker data
• Z-score SuStaIn (run_ZscoreSustain): For continuous biomarker measurements
• Mixture Model SuStaIn (run_MixtureSustain): For event-based disease progression modeling

Installation

Install the R package from GitHub:

# install.packages("devtools")
devtools::install_github("ucdavis/rSuStaIn")

Then install the pySuStaIn Python package:

library(rSuStaIn)
install_pySuStaIn()

Basic Usage

Ordinal SuStaIn (for scored data)

library(rSuStaIn)

# For ordinal/scored biomarker data
results <- run_OSA(
  patient_data = your_data,
  biomarker_levels = biomarker_levels,
  SuStaInLabels = names(biomarker_levels),
  N_startpoints = 25,
  N_S_max = 3
)

Z-score SuStaIn (for continuous data)

# For continuous biomarker measurements
results <- run_ZscoreSustain(
  data = your_biomarker_matrix,
  biomarker_labels = biomarker_names,
  N_startpoints = 25,
  N_S_max = 3
)

Mixture Model SuStaIn (event-based)

# For event-based disease progression modeling
results <- run_MixtureSustain(
  data = your_biomarker_matrix,
  biomarker_labels = biomarker_names,
  N_startpoints = 25,
  N_S_max = 3
)

Documentation

See the Getting Started vignette for detailed usage examples.

References

Young, A. L., Vogel, J. W., Aksman, L. M., Wijeratne, P. A., Eshaghi, A., Oxtoby, N. P., … & Alzheimer's Disease Neuroimaging Initiative. (2021). Ordinal SuStaIn: Subtype and Stage Inference for Clinical Scores, Visual Ratings, and Other Ordinal Data. Frontiers in Artificial Intelligence, 4, 613261.