: Model zoo for genomics
Model zoo for genomics
Kipoi (pronounce: kípi; from the Greek κήποι: gardens) is an API and a repository of ready-to-use trained models for genomics. It currently contains 2205 different models, covering canonical predictive tasks in transcriptional and post-transcriptional gene regulation. Kipoi's API is implemented as a python package (github.com/kipoi/kipoi) and it is also accessible from the command line or R.
Numbers
# of models: 2205
# of model groups: 37
# of contributors: 12
# of model groups supporting postprocessing:
- Variant effect prediction: 7/37
Model groups by framework
Model groups by tag
Getting Started
Use models
- Install Kipoi:
$ pip install kipoi - Choose the model from http://kipoi.org/models
- Use the model from CLI/python/bash by copying code snippets from http://kipoi.org/models/my_model
Contribute models
- Install Kipoi:
$ pip install kipoi - Run
$ kipoi ls - Explore the models locally at
~/.kipoi/models - Upload model weights to zenodo or figshare.
- Add the model to
~/.kipoi/models/(copy&edit the existing model or runkipoi init) - Test the model:
kipoi test ~/.kipoi/models/my_model - Submit the pull-request to github.com/kipoi/models
Contribute code
- Setup the development environment:
$ pip install kipoi[develop] - Read CONTRIBUTING.md
- Submit an issue/pull-request to github.com/kipoi/kipoi