Skip to content

Latest commit

 

History

History
34 lines (28 loc) · 934 Bytes

README.md

File metadata and controls

34 lines (28 loc) · 934 Bytes

ADNI

predict ADAS and MMSE score based on genotype

Dependencies

python packages

python3
numpy
tensorflow2
sklearn
pandas
matplotlib
rpy2

R package

glmnet

Data

Here is only the description, not public download. Because the access permission should apply from The Alzheimer's Disease Neuroimaging Initiative (ADNI, http://adni.loni.usc.edu).

The genetic data include:
(1) ADNI_JansenNG_33gene_unimputed.bed
(2) ADNI_JansenNG_33gene_unimputed.fam
(3) ADNI_JansenNG_33gene_unimputed.bim
The selected ADAS subjects and scores file which appear both in genetic and image data is:
(1) ADAS_selected.csv
The selected MMSE subjects and scores file is:
(2) MMSE_selected.csv

For more information about the scores, please check https://adni.loni.usc.edu/wp-content/uploads/2012/08/instruction-about-data.pdf

Methods

Lasso
CNN
CNN with pretrained representaiton by autoencoder