biosofa 0.7.1

Probabilistic factor analysis model with covariate guided factors

Semi-supervised Omics Factor Analysis (SOFA)

Introduction

Here we present semi-supervised probabilistic Factor Analysis (SOFA), a multi-omics integration method, which infers a set of low dimensional latent factors that represent the main sources of variability. SOFA enables the discovery of primary sources of variation while adjusting for known covariates and simultaneously disentangling variation that is shared between multiple omics modalities and specific to single modalities. The SOFA method is implemented in python using the Pyro framework for probabilistic programming.

Installation

To install SOFA first create Python 3.8 environment e.g. by

conda create --name sofa-env python=3.8
conda activate sofa-env

and install the package using

pip install biosofa

How to use SOFA for multi-omics analyses

A detailed manual with examples and how to use SOFA can be found here https://tcapraz.github.io/SOFA/index.html.

How to cite SOFA

Semi-supervised Omics Factor Analysis (SOFA) disentangles known sources of variation from latent factors in multi-omics data

Capraz, T., Vöhringer, H.S. and Huber, W.

bioRxiv 2024. doi: 10.1101/2024.10.10.617527.