CONCISE (COnvolutional Neural for CIS-regulatory Elements) is a model for predicting PTR features like mRNA half-life from cis-regulatory elements using deep learning.
Project description
<div align="center">
<img src="docs/img/concise_logo_text.jpg" alt="Concise logo" height="64" width="64">
</div>
# Concise: Keras extension for regulatory genomics
##
Concise (CONvolutional neural networks for CIS-regulatory Elements) is a Keras extension for regulatory genomics.
If allows you to:
1. pre-process sequence-related data (say convert a list of sequences into one-hot-encoded numpy arrays)
2. specify a keras model with additional utilites: concise provides custom `layers`, `initializers` and `regularizers` useful for regulatory genomics
3. tune the hyper-parameters (`hyopt`): concise provides convenience functions for working with `hyperopt` package.
4. interpret: concise layers contain visualization methods
5. share and re-use models: every concise component (layer, initializer, regularizer, loss) is fully compatible with keras:
- saving, loading and reusing the models works out-of-the-box
<!-- TODO - include image of concise -->
## Installation
Concise is available for python versions greater than 3.4 and can be installed from PyPI using `pip`:
```sh
pip install --process-dependency-links concise
```
`--process-dependency-links` is required in order to properly install the following github packages: [deeplift](https://github.com/kundajelab/deeplift) and [simdna](https://github.com/kundajelab/simdna/tarball/0.2#egg=simdna-0.2).
<!-- Make sure your keras is installed properly and configured with the backend of choice. -->
## Documentation
- <https://i12g-gagneurweb.in.tum.de/public/docs/concise/>
<img src="docs/img/concise_logo_text.jpg" alt="Concise logo" height="64" width="64">
</div>
# Concise: Keras extension for regulatory genomics
##
Concise (CONvolutional neural networks for CIS-regulatory Elements) is a Keras extension for regulatory genomics.
If allows you to:
1. pre-process sequence-related data (say convert a list of sequences into one-hot-encoded numpy arrays)
2. specify a keras model with additional utilites: concise provides custom `layers`, `initializers` and `regularizers` useful for regulatory genomics
3. tune the hyper-parameters (`hyopt`): concise provides convenience functions for working with `hyperopt` package.
4. interpret: concise layers contain visualization methods
5. share and re-use models: every concise component (layer, initializer, regularizer, loss) is fully compatible with keras:
- saving, loading and reusing the models works out-of-the-box
<!-- TODO - include image of concise -->
## Installation
Concise is available for python versions greater than 3.4 and can be installed from PyPI using `pip`:
```sh
pip install --process-dependency-links concise
```
`--process-dependency-links` is required in order to properly install the following github packages: [deeplift](https://github.com/kundajelab/deeplift) and [simdna](https://github.com/kundajelab/simdna/tarball/0.2#egg=simdna-0.2).
<!-- Make sure your keras is installed properly and configured with the backend of choice. -->
## Documentation
- <https://i12g-gagneurweb.in.tum.de/public/docs/concise/>
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
concise-0.6.0.tar.gz
(8.0 MB
view hashes)
Built Distributions
concise-0.6.0-py3.5.egg
(894.0 kB
view hashes)
concise-0.6.0-py2.py3-none-any.whl
(755.0 kB
view hashes)
Close
Hashes for concise-0.6.0-py2.py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 091f909644819ce664f0d374f0b1e5fddc143cdee93b7a738ec4107a0af16073 |
|
MD5 | 36712fdc9e40ae92024e78355e5959b0 |
|
BLAKE2b-256 | 0ef9e481040d047c8a0ab49af6a7b97e7b57a96e081d90c1b31b5c893a098bda |