Scalable Amortized GPLVMs for Single Cell Transcriptomics Data

TL;DR

This paper presents a scalable amortized Bayesian GPLVM tailored for single-cell RNA-seq data, improving interpretability and clustering over existing models by incorporating specialized design features.

Contribution

The authors introduce an improved amortized stochastic variational Bayesian GPLVM with specialized components, enhancing scalability and interpretability for single-cell transcriptomics.

Findings

Matches performance of scVI on synthetic and real datasets

Effectively incorporates cell-cycle and batch effects

Reveals interpretable latent structures in immunity data

Abstract

Dimensionality reduction is crucial for analyzing large-scale single-cell RNA-seq data. Gaussian Process Latent Variable Models (GPLVMs) offer an interpretable dimensionality reduction method, but current scalable models lack effectiveness in clustering cell types. We introduce an improved model, the amortized stochastic variational Bayesian GPLVM (BGPLVM), tailored for single-cell RNA-seq with specialized encoder, kernel, and likelihood designs. This model matches the performance of the leading single-cell variational inference (scVI) approach on synthetic and real-world COVID datasets and effectively incorporates cell-cycle and batch information to reveal more interpretable latent structures as we demonstrate on an innate immunity dataset.