Machine learning applications for dadi
Project description
Machine Learning Applications for Diffusion Approximation of Demographic Inference
Introduction
Diffusion Approximation of Demographic Inference (dadi) is a powerful software tool for simulating the joint frequency spectrum (FS) of genetic variation among multiple populations and employing the FS for population-genetic inference. Here we introduce machine learning-based tools for easier application of dadi's underlying demographic models. These machine learning models were trained on dadi-simulated data and can be used to make quick predictions on dadi demographic model parameters given FS input data from user and specified demographic model. The pipeline we used to train the machine learning models are also available here for users interested in using the same framework to train a new predictor for their customized demographic models.
Installation
Get the dadi-ml repo
Clone this repo to your local directory and cd
into the dadi-ml
dir
$ git clone https://github.com/lntran26/dadi-ml.git
$ cd dadi-ml/
Set up your python environment to run the dadi-ml pipeline
We recommend you start by creating a new conda
environment. This can be done using the command below, which will create a new conda
env called dadi-ml
and installed the required packages to this env. The env can then be activated for each subsequent use.
$ conda env create -f environment.yml
$ conda activate dadi-ml
Requirements
- Python 3.9+
- dadi
- scikit-learn 1.0.2
- MAPIE 0.3.1
References
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
Built Distribution
Hashes for project_ml_ab3nzac1yc_hatch-1.0.1.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | 3e1cc10a4992b90cf5a887c9c13dc3d7b51d42ee7bed02283d6aaecc57936bfa |
|
MD5 | d1f4e2ce48754b2e6a1c3b20d3a51f98 |
|
BLAKE2b-256 | c682c5d7159d5f221d094fc1d294f21848c9eace6fdf1cf34a92aacb041227e2 |
Hashes for project_ml_ab3nzac1yc_hatch-1.0.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 30311a49c3cac42f654e88d34595df56863fb7998eb6fefdcb0c4d7c5e5e61ac |
|
MD5 | 2a4c264d7ce379f0249bf1e0289c01dd |
|
BLAKE2b-256 | 32586c16fcb09c6f4d38b936591b51b02a47ea37595c3015fb98045c7d6114df |