GenoML developers advocate open science. The following are the underlying principles of GenoML development:
Little to learn - The goal of GenoML is to democratize complex genomics and machine learning workflows. Thoughtfully designed for newcomers, if a user can
ls, they should be able to use GenoML.
Intuitive - Everything has to be simple, straightforward, and effective, from data munging to a tuned model in a few lines of code.
Layered architecture - GenoML is more than a tool; it is an ecosystem that will continuously grow, experimenting with new ideas and innovations. Workflows are kept in logical layers; to change or update one module and not affect the others.
Intelligent defaults - Systematic research is done to set optimized defaults for varying inputs. The default settings are sensible and validated for most workflows to keep modules un-cluttered and to run smoothly. At the same time, providing manual options for advanced users.
No vendor lock-in - Integration with other code, products, and platforms should be hassle-free. GenoML is open source and will remain free and public under the Apache 2.0 license.