Robust Multi-view Co-expression Network Inference

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Pandeva, Teodora, Jonker, Martijs, Hamoen, Leendert, Mooij, Joris, Forré, Patrick
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910625069793280
author Pandeva, Teodora
Jonker, Martijs
Hamoen, Leendert
Mooij, Joris
Forré, Patrick
author_facet Pandeva, Teodora
Jonker, Martijs
Hamoen, Leendert
Mooij, Joris
Forré, Patrick
contents Unraveling the co-expression of genes across studies enhances the understanding of cellular processes. Inferring gene co-expression networks from transcriptome data presents many challenges, including spurious gene correlations, sample correlations, and batch effects. To address these complexities, we introduce a robust method for high-dimensional graph inference from multiple independent studies. We base our approach on the premise that each dataset is essentially a noisy linear mixture of gene loadings that follow a multivariate $t$-distribution with a sparse precision matrix, which is shared across studies. This allows us to show that we can identify the co-expression matrix up to a scaling factor among other model parameters. Our method employs an Expectation-Maximization procedure for parameter estimation. Empirical evaluation on synthetic and gene expression data demonstrates our method's improved ability to learn the underlying graph structure compared to baseline methods.
format Preprint
id arxiv_https___arxiv_org_abs_2409_19991
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Robust Multi-view Co-expression Network Inference
Pandeva, Teodora
Jonker, Martijs
Hamoen, Leendert
Mooij, Joris
Forré, Patrick
Machine Learning
Quantitative Methods
Applications
Unraveling the co-expression of genes across studies enhances the understanding of cellular processes. Inferring gene co-expression networks from transcriptome data presents many challenges, including spurious gene correlations, sample correlations, and batch effects. To address these complexities, we introduce a robust method for high-dimensional graph inference from multiple independent studies. We base our approach on the premise that each dataset is essentially a noisy linear mixture of gene loadings that follow a multivariate $t$-distribution with a sparse precision matrix, which is shared across studies. This allows us to show that we can identify the co-expression matrix up to a scaling factor among other model parameters. Our method employs an Expectation-Maximization procedure for parameter estimation. Empirical evaluation on synthetic and gene expression data demonstrates our method's improved ability to learn the underlying graph structure compared to baseline methods.
title Robust Multi-view Co-expression Network Inference
topic Machine Learning
Quantitative Methods
Applications
url https://arxiv.org/abs/2409.19991