lantern
lantern copied to clipboard
Interpretable genotype-phenotype landscape modeling
#+TITLE: LANTERN: an interpretable genotype-phenotype landscape model
DISCLAIMER: LANTERN is currently in alpha and the primary interface is subject to change
-
Documentation The official documentation with usage is available at: [[https://lantern-gpl.readthedocs.io/en/latest/index.html][https://lantern-gpl.readthedocs.io]]. Documentation covers instructions for running ~lantern~ on GPL data, including a demonstration on an example dataset. On a typical desktop, examples should run in under ten minutes.
-
System Requirements
** Hardware requirements ~lantern~ requires only a standard computer with enough RAM to support the in-memory operations. Runtime of model training on large-scale GPL datasets ($n \geq 10,000$) will benefit from GPU hardware but is not strictly necessary.
** Software requirements This package is supported for macOS and Linux. The package has been tested on the following systems:
- macOS: Catalina (10.15)
- Linux: Ubuntu 18.04
** Python Dependencies ~lantern~ depends primarily on ~pytorch~ and ~gpytorch~, as well as the components of the Python scientific stack:
- ~pandas~
- ~numpy~
- Installation guide Installation time for ~lantern~ is typically less than one minute.
** Install from github ~LANTERN~ currently must be installed from this repository directly, either through pip: #+begin_src bash python -m pip install git+https://github.com/usnistgov/lantern.git #+end_src of by cloning this repository: #+begin_src bash git clone https://github.com/usnistgov/lantern.git cd lantern python setup.py install #+end_src In either case, it is recommended to install inside a [[https://docs.python.org/3/tutorial/venv.html][virtual environment]].
-
Manuscript reproduction Source code to reproduce the analysis of the LANTERN manuscript are available at [[https://github.com/usnistgov/lantern/tree/master/manuscript][github.com/usnistgov/lantern/tree/master/manuscript]].
-
License This project is covered under the NIST Software License