dsaffo / genoml

Core pipeline of GenoML

GenoML-core

GenoML is an Automated Machine Learning (AutoML) for Genomic. This is the core package of GenoML. this repo is under development, please report any issues (bug, performance, documentation) on the GenoML issues page.

Here are some quick "get started" exmaples, please checkout the additional options and details in the Usage and CLI. In general, use linux or mac with python > 3.5 for best results.

Install

Run:

pip install genoml

Train the ML model

You can use the IPDGC (International Parkinson's Disease Genomics Consortium) test data. This data is a simulation of the genetic and clinical data used for Parkinson's diagnosis in previous publications. You can find it at IPDGC example data.

Download and unzip data:

wget https://github.com/ipdgc/GenoML-Brief-Intro/raw/master/exampleData.zip
unzip exampleData.zip 

To train, run:

genoml-train --geno-prefix=./exampleData/training --pheno-file=./exampleData/training.pheno --model-dir=./exampleModel

Final tuned model and performance metrics are stored in the --model-dir directory.

Using the trained ML model for inference

genoml-inference --model-dir=./exampleModel --valid-dir=./exampleData --valid-geno-prefix=./exampleData/validation --valid-pheno-file=./exampleData/validation.pheno

Valdiation results and model performance metrics are stored in the --valid-dir directory.

For debugging purposes, include the -v or -vvv flags at the end of a command.

Report issues

Please report any issue or suggestions on the GenoML issues page.