Skip to content

README.md

Latest commit

 

History

History
164 lines (131 loc) · 4.62 KB

README.md

File metadata and controls

164 lines (131 loc) · 4.62 KB
Raw

Soybean-CNN

Research about the Soybean CNN for the 2025 BIBM Conference.

We applied a saliency map approach to measure phenotype contribution for genome wide association study. Creates output files on average saliency graph along with top SNPs with highest average saliency. On program completion, it provides the average PCC (Pearson correlation coefficient) along with a prompt for the user to search saliency based on SNP name. The program is implemented using Keras3.5 and Tensorflow backend with python 3.1

Link to the published paper: https://ieeexplore.ieee.org/document/11356176

Prerequisites

Python packages required:

absl_py==2.1.0 
astunparse==1.6.3 
certifi==2025.1.31 
charset_normalizer==3.4.1 
contourpy==1.3.1 
cycler==0.12.1 
flatbuffers==25.1.24 
fonttools==4.56.0 
gast==0.6.0 
google-pasta==0.2.0 
grpcio==1.67.0 
h5py==3.12.0 
idna==3.10 
joblib==1.4.2 
keras==3.5.0 
kiwisolver==1.4.8 
libclang==14.0.1 
Markdown==3.7 
markdown_it_py==3.0.0 
MarkupSafe==2.1.5 
matplotlib==3.10.0 
mdurl==0.1.2 
ml_dtypes==0.4.0 
namex==0.0.8 
np_utils==0.6.0
numpy==1.26.4 
opt_einsum==3.4.0 
optree==0.12.1 
packaging==24.2 
pandas==2.2.3 
pillow==11.1.0 
protobuf==4.25.4 
pygments==2.19.1 
pyparsing==3.2.3
python_dateutil==2.9.0.post0 
pytz==2025.2
requests==2.32.3 
rich==13.9.4 
scikit_learn==1.5.2 
scipy==1.15.1 
six==1.17.0 
sklearn==0.0 
tensorboard==2.17.1 
tensorboard_data_server==0.7.2 
tensorflow==2.17.0 
tensorflow_io_gcs_filesystem==0.32.0 
termcolor==2.5.0 
threadpoolctl==3.6.0 
typing_extensions==4.12.2 
tzdata==2025.2
urllib3==2.3.0 
werkzeug==3.1.3 
wrapt==1.17.2

Running the program

The scripts train and test model with 10 fold cross validation and plot a comparison of genotype contribution using saliency map value and Wald test value.

  • height.py/moisture.py/oil.py/protein.py - Executive scripts.
  • run_all_folds.py - Runs all 10 folds individually.
  • summarize_folds.py - Summarizes all 10 folds without re-running program.
  • requirements.txt - Requirements of Python modules for program.
  • IMP_height.txt/IMP_moisture.txt/IMP_oil.txt/IMP_protein.txt - Inputs of imputed genotype matrix.
  • QA_height.txt/QA_moisture.txt/QA_oil.txt/QA_protein.txt - Inputs of quality assured non-imputed genotype matrix.
  • polytest.txt - Genotype contribution using Wald test score. Run with SoyNAM R package.
  • saliency_value.txt - Genotype contribution calculated using saliency map approach.

In all three python files, change the global variable NUM_FOLDS to the folds you wish to run.

Running on Digital Alliance Servers

After SSH'ing into servers, preform the following to run the CNN for the height phenotype. (Bash file, CNN_setup.sh, provided to install packages)

module load python/3.10
virtualenv --no-download ENV
source ENV/bin/activate
cd HEIGHT
python3 height.py

alternative:
python3 height.py --fold 3 (Runs individual fold)

Summarize after 10 folds completed:

python3 summarize_folds.py

alternative:

python3 height.py --summary

The user can change the phenotype from height by changing the directory and program name. For example, run the CNN for the moisture phenotype, substitute

cd HEIGHT
python3 height.py

with

cd MOISTURE
python3 moisture.py

Running on Docker

Build image (requirements.txt file provides to install needed Python packages):

docker build -t snp-gwas-predictor .

Run full program:

Linux/macOS: docker run --rm -it -v "$(pwd):/app" snp-gwas-predictor
Windows CMD: docker run --rm -it -v "%cd%:/app" snp-gwas-predictor
Windows PowerShell: docker run --rm -it -v "${PWD}:/app" snp-gwas-predictor

alternatives (height.py can be replaced with any other phenotype file):

docker run --rm -v "%cd%:/app" snp-gwas-predictor python3 height.py --fold 2 (Runs selected fold and appends PCC values to export file)
docker run --rm snp-gwas-predictor python3 height.py --fold 2 (Won’t save to PCC export file)

Summarize after 10 folds completed (height.py can be replaced with any other phenotype file):

Linux/macOS: docker run --rm -it -v "$(pwd):/app" snp-gwas-predictor python3 summarize_folds.py
Windows CMD: docker run --rm -it -v "%cd%:/app" snp-gwas-predictor python3 summarize_folds.py 
Windows PowerShell: docker run --rm -it -v "${PWD}:/app" snp-gwas-predictor python3 summarize_folds.py

alternative:

docker run --rm -it -v "%cd%:/app" snp-gwas-predictor python3 height.py --summary

Current Authors

Jake Goode, Madhurika Madhu, Raeein Bagheri, and Yan Yan

School of Computer Science

University of Guelph, Guelph ON, Canada

Previous Authors

Yang Liu

University of Missouri, Columbia MO, USA

License

GNU v2.0