Forecasting individual progression trajectories in Alzheimer's disease – software and source data

# Content

The repository contains the frozen version of the Python libraries that were used to generate results of the article Forecasting individual progression trajectories in Alzheimer’s disease, Maheux E., et al. (2022)

It contains two zipped Python packages:

• leaspy for training and predicting progression trajectories based on AD Course Map, Linear mixed model and No-change prediction models
• rnn-ad for training and predicting progression trajectories based on RNN-AD model

Read-me and licence files are provided within each package.

It also contains the source_data.xslx that were used to generate all figures in the main manuscript, as well as in the supplementary information.

# Data availability

Please note that data and trained models are **NOT** provided, due to confidentiality issues.

• The ADNI and AIBL data used in this study are available in the database of the laboratory of neuroimaging at the university of Southern California under accession code at http://adni.loni.usc.edu.
• The J-ADNI data used in this study are available in the NBDC Human Database under accession code at http://humandbs.biosciencedbc.jp/en/.
• The PharmaCog data used in this study are available in the NeuGRID2 platform under access code at https://www.neugrid2.eu/ (https://doi.org/10.17616/R31NJN1E)
• The MEMENTO data used in this study are available in Dementia Platform UK under accession code at https://portal.dementiasplatform.uk/CohortDirectory/Item?fingerPrintID=MEMENTO

# Bugs and questions

In case you have any question regarding the software or the source data, please contact the corresponding author of the article.