Metadata-Version: 2.4
Name: dlim
Version: 1.1.0
Summary: Direct Latent Interpretable Model (D-LIM): An interpretable neural network for mapping genotype to fitness.
Author: Shuhui Wang, Alexandre Allauzen, Philippe Nghe, Vaitea Opuu
Author-email: vaiteaopuu@gmail.com
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.7
Description-Content-Type: text/markdown
License-File: LICENSE.txt
Requires-Dist: numpy>=1.21.0
Requires-Dist: torch>=1.10.0
Dynamic: author
Dynamic: author-email
Dynamic: classifier
Dynamic: description
Dynamic: description-content-type
Dynamic: license-file
Dynamic: requires-dist
Dynamic: requires-python
Dynamic: summary

# D-LIM (Direct-Latent Interpretable Model)

# Overview

D-LIM (Direct-Latent Interpretable Model) is a neural network that enhances
genotype-fitness mapping by combining interpretability with predictive accuracy.
It assumes independent phenotypic influences of genes on fitness, leading to
advanced accuracy and insights into phenotype analysis and epistasis. The model
includes an extrapolation method for better understanding of genetic
interactions and integrates multiple data sources to improve performance in
low-data biological research.


# System Requirements

## Hardware requirements
   **D-LIM** requires only a standard computer with enough RAM to
   support the in-memory operations. 

## Software requirements
   This package is supported for Linux. The package has been tested on the following systems:
   - Linux: Ubuntu 20.04

## Python Dependencies
   **D-LIM** depends primarily on **pytorch**, as well as
   the components of the Python scientific stack:
   - `pandas`
   - `numpy`

## Installation guide
- Install the package from Pypi:
```
pip install dlim
```

Or install it from the sources:
```
git clone https://github.com/LBiophyEvo/D-LIM-model.git
cd D-LIM-model 
pip install -e .
```

## Documentation 
The official documentation with usage is available at: https://d-lim.readthedocs.io/en/latest/. Documentation covers instructions for running D-LIM on simulated and experimental data, including demonstrations on some simple datasets. 

# Data avaibility 
- Simulated data: see `src_simulate_data\`
- Experimental data:
   - Data from [Kemble et al.](https://www.science.org/doi/10.1126/sciadv.abb2236) can be found in `data`. 
   - Data of protein-protein interaction is from [Diss et al.](https://elifesciences.org/articles/32472).
   - Data of yeast with different mutations and enviroments is from [Kinsler et al.](https://elifesciences.org/articles/61271). 
You can download data from our Google driver link: [Data used in D-LIM](https://drive.google.com/drive/folders/1hwixojm3thyYpf8X6qPG7NIvxQseFDKz?usp=sharing).

# Manuscript reproduction
  Source code to reproduce the analysis of the **D-LIM** manuscript are
  available at `reproducibility` folder in [D-LIM GitHub](https://github.com/LBiophyEvo/D-LIM-model.git).
  
# License
  This project is covered under the *MIT License*
