VEGA is an interpretable generative model for inferring biological network activity in single-cell transcriptomics

Seninge, Lucas; Anastopoulos, Ioannis N.; Ding, Hongxu; Stuart, Joshua M.

doi:10.1038/s41467-021-26017-0

Cited by 64 publications

(66 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We first proposed a regularized linear decoder to include domain knowledge into autoencoders for single-cell data at a conference 86 , with scalable and expressive embeddings when compared to existing factor models, such as f-scLVM 87 . Recent approaches such as VEGA 88 , scTEM 89 and pmVAE 90 also feature VAE-based architectures with linear decoders or training separate VAEs for each GP yet connected via a global loss in the case of pmVAE. In contrast, expiMap aims toward interpretable reference mapping allowing to fuse reference atlases with GPs and enabling the query of genes or GPs.…”

Section: Discussionmentioning

confidence: 99%

Biologically informed deep learning to infer gene program activity in single cells

Lotfollahi

Rybakov

Hrovatin

et al. 2022

Preprint

View full text Add to dashboard Cite

The increasing availability of large-scale single-cell datasets has enabled the detailed description of cell states across multiple biological conditions and perturbations. In parallel, recent advances in unsupervised machine learning, particularly in transfer learning, have enabled fast and scalable mapping of these new single-cell datasets onto reference atlases. The resulting large-scale machine learning models however often have millions of parameters, rendering interpretation of the newly mapped datasets challenging. Here, we propose expiMap, a deep learning model that enables interpretable reference mapping using biologically understandable entities, such as curated sets of genes and gene programs. The key concept is the substitution of the uninterpretable nodes in an autoencoder's bottleneck by labeled nodes mapping to interpretable lists of genes, such as gene ontologies, biological pathways, or curated gene sets, for which activities are learned as constraints during reconstruction. This is enabled by the incorporation of predefined gene programs into the reference model, and at the same time allowing the model to learn de novo new programs and refine existing programs durin reference mapping. We show that the model retains similar integration performance as existing methods while providing a biologically interpretable framework for understanding cellular behavior. We demonstrate the capabilities of expiMap by applying it to 15 datasets encompassing five different tissues and species. The interpretable nature of the mapping revealed unreported associations between interferon signaling via the RIG-I/MDA5 and GPCRs pathways, with differential behavior in CD8+ T cells and CD14+ monocytes in severe COVID-19, as well as the role of annexins in the cellular communications between lymphoid and myeloid compartments for explaining patient response to the applied drugs. Finally, expiMap enabled the direct comparison of a diverse set of pancreatic beta cells from multiple studies where we observed a strong, previously unreported correlation between the unfolded protein response and asparagine N-linked glycosylation. Altogether, expiMap enables the interpretable mapping of single cell transcriptome data sets across cohorts, disease states and other perturbations.

show abstract

Section: Discussionmentioning

confidence: 99%

Biologically informed deep learning to infer gene program activity in single cells

Lotfollahi

Rybakov

Hrovatin

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Over the past years, deep learning (DL) has become an essential tool for analysis (Lopez et al, 2020) and interpretation (Rybakov et al, 2020) of scRNA-seq data. Representation learning in particular, has been useful not only for identifying cellular heterogeneity and integration , or mapping query to reference datasets (Lotfollahi et al, 2022), but also in the context of modelling single-cell perturbation responses (Rampášek et al, 2019;Seninge et al, 2021;Lotfollahi et al, 2019;.…”

Section: Related Workmentioning

confidence: 99%

Predicting single-cell perturbation responses for unseen drugs

Hetzel¹,

Böhm²,

Kilbertus³

et al. 2022

Preprint

View full text Add to dashboard Cite

Single-cell transcriptomics enabled the study of cellular heterogeneity in response to perturbations at the resolution of individual cells. However, scaling highthroughput screens (HTSs) to measure cellular responses for many drugs remains a challenge due to technical limitations and, more importantly, the cost of such multiplexed experiments. Thus, transferring information from routinely performed bulk RNA-seq HTS is required to enrich single-cell data meaningfully. We introduce a new encoder-decoder architecture to study the perturbational effects of unseen drugs. We combine the model with a transfer learning scheme and demonstrate how training on existing bulk RNA-seq HTS datasets can improve generalisation performance. Better generalisation reduces the need for extensive and costly screens at single-cell resolution. We envision that our proposed method will facilitate more efficient experiment designs through its ability to generate insilico hypotheses, ultimately accelerating targeted drug discovery.

show abstract

“…While this approach does not require labeled data, it can only be applied to VAE models, and can not be applied to standard, deterministic autoencoders. Another fully unsupervised approach to ranking important pathways examines the L2 norm of the weights connecting each latent pathway to the reconstruction output; however, this metric is obviously limited to models with linear decoders [21]. Unlike all of these approaches, our proposed pathway attribution (see Methods) is both fully unsupervised, meaning it requires no labeled data, and model agnostic, meaning it can be applied to any model regardless of architecture or implementation details.…”

Section: Additional Covariatesmentioning

confidence: 99%

“…These biologically-constrained networks have been used in a supervised setting to improve the prediction of survival or treatment resistance from cancer gene expression or mutational status [17,18], and to improve Genome Wide Association studies by aggregating the effects of single nucleotide polymorphisms (SNPs) into SNP sets [19]. In particular, a variety of recent works have proposed using biologicaly-constrained autoencoders to model gene expression data [20][21][22].…”

Section: Introductionmentioning

confidence: 99%

Principled feature attribution for unsupervised gene expression analysis

Janizek

Spiro

Çelik

et al. 2022

Preprint

View full text Add to dashboard Cite

As interest in unsupervised deep learning models for the analysis of gene expression data has grown, an increasing number of methods have been developed to make these deep learning models more interpretable. These methods can be separated into two groups: (1) post hoc analyses of black box models through feature attribution methods and (2) approaches to build inherently interpretable models through biologically-constrained architectures. In this work, we argue that these approaches are not mutually exclusive, but can in fact be usefully combined. We propose a novel unsupervised pathway attribution method, which better identifies major sources of transcriptomic variation than prior methods when combined with biologically-constrained neural network models. We demonstrate how principled feature attributions aid in the analysis of a variety of single cell datasets. Finally, we apply our approach to a large dataset of post-mortem brain samples from patients with Alzheimer's disease, and show that it identifies Mitochondrial Respiratory Complex I as an important factor in this disease.

show abstract

VEGA is an interpretable generative model for inferring biological network activity in single-cell transcriptomics

Cited by 64 publications

References 40 publications

Biologically informed deep learning to infer gene program activity in single cells

Biologically informed deep learning to infer gene program activity in single cells

Predicting single-cell perturbation responses for unseen drugs

Principled feature attribution for unsupervised gene expression analysis

Contact Info

Product

Resources

About