README

# Operating systems: Ubuntu
# Software: Python(3.6.8)
# Required packages: argparse(1.1), numpy(1.15.4), keras(2.2.4), PIL(5.4.1), scipy(1.2.0), matplotlib(3.0.2) and vis.visualization
# No non-standard hardware is required
# Ready to use. No need to install.

# Use fundus image to classify whether the late AMD progression time is before or after 3 years 
python code_img2amdState2time.py --cutoff_yr 3 --input_image test1.jpg --out_file output.txt --weights "/directory/weights"

# To generate saliency map to visualize what part of the image is focused by CNN to make the classification decision
python code_img2amdState2time_saliency.py --cutoff_yr 3 --input_image test1.jpg --out_image output.png --weights "/directory/weights"

# Use fundus image and 52 reported SNPs to predict whether the late AMD progression time is before or after 3 years
python code_imgNgeno2amdState2time.py --cutoff_yr 3 --input_image test1.jpg --input_geno input.geno --out_file output.txt --weights "/directory/weights"

# Expected run time should be seconds or up to a few minutes.