Brain AutoEncoder CNN Project
Brain Autoencoder for Convolutional Neural Networks (CNN)
Results for phenotype discovery by unsupervised deep representation learning for genetic association studies of brain imaging.
Projects
DeepEndo project resources.
Project pages and downloadable resources for unsupervised deep-learning-derived imaging phenotypes and genome-wide association studies.
DeepEndo
Imaging phenotypes for genetic discovery
DeepEndo focuses on phenotype discovery by unsupervised deep representation learning for genetic association studies of brain imaging. The resources below mirror the AIGI project content and point to papers, code, summary statistics, enrichment results, and visualization outputs.
JAGWAS Project
Joint Analysis of multi-phenotype GWAS (JAGWAS)
Efficient multi-phenotype genome-wide analysis identifies genetic associations for unsupervised deep-learning-derived high-dimensional brain imaging phenotypes.
UDIP-FA Project
Unveiling genetic architecture of white matter microstructure through unsupervised deep representation learning of fractional anisotropy images
UDIP-FA applies unsupervised deep representation learning to diffusion MRI fractional anisotropy images to derive high-dimensional white matter phenotypes for imaging genetics. The linked viewer provides interactive visualization for 128 learned dimensions and tract-level statistics.