Making Multi-Axis Models Robust to Multiplicative Noise: How, and Why?

TL;DR

This paper introduces MED-MAGMA, a graph-learning algorithm designed to robustly fit multi-axis models affected by multiplicative noise, common in biological data like RNA sequencing.

Contribution

The paper presents a novel algorithm, MED-MAGMA, that effectively handles multiplicative noise in multi-axis models, improving network learning in noisy biological datasets.

Findings

MED-MAGMA outperforms prior methods on all tested datasets.

The algorithm learns networks with better local and global structure.

MED-MAGMA is available as an open-source Python package.

Abstract

In this paper we develop a graph-learning algorithm, MED-MAGMA, to fit multi-axis (Kronecker-sum-structured) models corrupted by multiplicative noise. This type of noise is natural in many application domains, such as that of single-cell RNA sequencing, in which it naturally captures technical biases of RNA sequencing platforms. Our work is evaluated against prior work on each and every public dataset in the Single Cell Expression Atlas under a certain size, demonstrating that our methodology learns networks with better local and global structure. MED-MAGMA is made available as a Python package (MED-MAGMA).