Quantitative assessment of protein activity in orphan tissues and single cells using the metaVIPER algorithm

Abstract

We and others have shown that transition and maintenance of biological states is controlled by master regulator proteins, which can be inferred by interrogating tissue-specific regulatory models (interactomes) with transcriptional signatures, using the VIPER algorithm. Yet, some tissues may lack molecular profiles necessary for interactome inference (orphan tissues), or, as for single cells isolated from heterogeneous samples, their tissue context may be undetermined. To address this problem, we introduce metaVIPER, an algorithm designed to assess protein activity in tissue-independent fashion by integrative analysis of multiple, non-tissue-matched interactomes. This assumes that transcriptional targets of each protein will be recapitulated by one or more available interactomes. We confirm the algorithm’s value in assessing protein dysregulation induced by somatic mutations, as well as in assessing protein activity in orphan tissues and, most critically, in single cells, thus allowing transformation of noisy and potentially biased RNA-Seq signatures into reproducible protein-activity signatures.

AFF3 and BACH2 are master regulators of metabolic inflexibility, β/α-cell transition, and dedifferentiation in type 2 diabetes

“…2B), without noticeable donor-to-donor variation. Consistent with prior results 27 , this shows that protein-activity-based cluster analyses can buffer technical bias and batch effects typical of scRNA-seq profiles to reveal hidden, biologically-relevant population structures.…”

“…B (illustrated in a REVIGO summary in Supplementary Fig. 4) 24, 25, 27 , shows highly significant enrichment for GO terms such as metabolism, development, and differentiation in cluster A. Among the proteins subsumed under metabolism in GO, we detected a striking activation of PPARα and PPAR γ (Fig.…”

“…1A) trumped cellular identities, as indicated by the emergence of single-cell clusters that were largely independent of disease status. To buffer donor-to-donor variation and batch effect, we thus relied on the network-based metaVIPER algorithm, which has been shown to be highly effective in removing systematic sample-to-sample bias and batch effects that do not reflect the biology of the cellular state 27 .…”

“…As a result, reliable differential gene expression assessment is challenging. As previously shown 27 , metaVIPER is especially well suited to overcome this limitation because the activity of a protein is measured from tens to hundreds of its transcriptional targets, thus allowing reliable protein activity assessment from profiles with only 20K to 50K reads. Indeed, Spearman correlation between a 30M and a 50K read, VIPER-inferred protein activity profile is ρ = 0.85, while for gene expression it is only ρ = 0.3 25 .…”

“…Using single-cell RNA-seq (scRNA-seq) from human islets—harvested from either normal (ND) or T2D donors—we first generated islet-cell-specific transcriptional regulatory networks that were then used to measure the transcriptional activity of ∼5600 transcriptional regulator proteins—including transcription factors, co-factors, and signal-transduction proteins— using an extension of the VIPER 25 algorithm for single-cell analysis (metaVIPER 27 ). Basically, the activity of each protein is measured based on the expression of its transcriptional targets, akin to a multiplexed gene-reporter assay.…”