Comprehensive single cell mRNA profiling reveals a detailed roadmap for pancreatic endocrinogenesis

The introduction of RNA velocity in single cells has opened up new ways of studying cellular differentiation. The originally proposed framework obtains velocities as the deviation of the observed ratio of spliced and unspliced mRNA from an inferred steady state. Errors in velocity estimates arise if the central assumptions of a common splicing rate and the observation of the full splicing dynamics with steady-state mRNA levels are violated. With scVelo (https://scvelo.org), we address these restrictions by solving the full transcriptional dynamics of splicing kinetics using a likelihood-based dynamical model. This generalizes RNA velocity to a wide variety of systems comprising transient cell states, which are common in development and in response to perturbations. We infer gene-specific rates of transcription, splicing and degradation, and recover the latent time of the underlying cellular processes. This latent time represents the cell's internal clock and is based only on its transcriptional dynamics. Moreover, scVelo allows us to identify regimes of regulatory changes such as stages of cell fate commitment and, therein, systematically detects putative driver genes. We demonstrate that scVelo enables disentangling heterogeneous subpopulation kinetics with unprecedented resolution in hippocampal dentate gyrus neurogenesis and pancreatic endocrinogenesis. We anticipate that scVelo will greatly facilitate the study of lineage decisions, gene regulation, and pathway activity identification.Contributions VB designed and developed the method, implemented scVelo and analyzed the data. FJT conceived the study with contributions from VB and FAW. VB, FAW and FJT wrote the manuscript with contributions from the coauthors. SP contributed to developing scVelo, and ML contributed to developing validation metrics. All authors read and approved the final manuscript.

show abstract

“…6). lineage, the endocrine fates could not be clearly delineated, and incongruous subpopulation flows emerged 20 .…”

Section: Resolving the Heterogeneous Population Kinetics In Dentate Gmentioning

confidence: 99%

“…Next, we demonstrate scVelo's capabilities to delineate transient lineages in endocrine development in the mouse pancreas, with transcriptome profiles sampled from E15.520 . Endocrine cells are derived from endocrine progenitors located in the pancreatic epithelium, marked by transient expression of the transcription factor Ngn3.…”

mentioning

confidence: 99%

Generalizing RNA velocity to transient cell states through dynamical modeling

Bergen

Lange

Peidli

et al. 2019

Preprint

253

349

View full text Add to dashboard Cite

show abstract

“…The Pancreas data set (Bastidas-Ponce et al 2019) stems from a study of endocrine development in mouse, and was acquired with the 10x Genomics Chromium platform, using v2 chemistry. We downloaded the FASTQ files containing the reads from the cells at stage E15.5 from the Gene Expression Omnibus, accession number GSE132188.…”

Section: Methodsmentioning

confidence: 99%

Preprocessing choices affect RNA velocity results for droplet scRNA-seq data

Soneson

Srivastava

Patro

et al. 2020

Preprint

View full text Add to dashboard Cite

12Experimental single-cell approaches are becoming widely used for many purposes, including inves-13 tigation of the dynamic behaviour of developing biological systems. Consequently, a large number 14 of computational methods for extracting dynamic information from such data have been developed. 15 One example is RNA velocity analysis, in which spliced and unspliced RNA abundances are jointly 16 modeled in order to infer a 'direction of change' and thereby a future state for each cell in the gene 17 expression space. 18 Naturally, the accuracy and interpretability of the inferred RNA velocities depend crucially on the 19 correctness of the estimated abundances. Here, we systematically compare four widely used quan-20 tification tools, in total yielding twelve different quantification approaches, in terms of their estimates 21 of spliced and unspliced RNA abundances in four experimental droplet scRNA-seq data sets. We 22show that there are substantial differences between the quantifications obtained from different tools, 23 and identify typical genes for which such discrepancies are observed. We further show that these 24 abundance differences propagate to the downstream analysis, and can have a large effect on estimated 25 velocities as well as the biological interpretation. 26Our results highlight that abundance quantification is a crucial aspect of the RNA velocity anal-27 ysis workflow, and that both the definition of the genomic features of interest and the quantification 28 algorithm itself require careful consideration. 29 Single-cell RNA-seq (scRNA-seq) enables high-throughput profiling of gene expression on a transcriptome-31 wide scale in individual cells. The increased resolution compared to bulk RNA-seq, where only average 32 expression profiles of populations of cells are obtained, provides vastly improved potential to study 33 a variety of biological questions. One such question concerns the dynamics of biological systems, re-34 flected in, e.g., cellular differentiation and development. While such dynamical processes would ideally 35 be studied via repeated transcriptome-wide expression profiling of the same cells over time, this is not 36 possible with current scRNA-seq protocols. Existing analysis methods for so called trajectory inference 37 are instead applied to one or several snapshots of a population of cells, assumed to comprise all stages 38 of the trajectory of interest. Many computational methods for trajectory inference from scRNA-seq have 39 been presented in the literature (reviewed by Saelens et al. (2019)). These methods typically use the 40 similarity of the gene expression profiles between cells to construct a (possibly branching) path through 41 the observed set of cells, representing the trajectory of interest. Projecting the cells onto this path then 42 provides an ordering of the cells by so called pseudotime. 43 A different approach to the investigation of developmental processes in scRNA-seq data instead 44 exploits the underlying molecular dynamics. The feasibility of ...

show abstract

“…We obtained the pseudo-time ordering by projecting the member cells onto the principal curve (Hastie and Stuetzle 1989). To confirm that VeTra can cluster multifurcated trajectories, we applied VeTra to two scRNAseq datasets for the pancreatic development (Bastidas-Ponce et al 2019;Bergen et al 2020) and neural lineages in the hippocampus (La Manno et al 2018). From the transcriptome of pancreatic development, VeTra identified three trajectory clusters of i) alpha cell differentiation from endocrine progenitors (EP), ii) beta/epsilon differentiation from EP, and iii) ductal cells ( Figure 1D).…”

Section: Methodsmentioning

confidence: 99%

VeTra: a tool for trajectory inference based on RNA velocity

Weng

Kim

Won

2020

Preprint

View full text Add to dashboard Cite

MotivationTrajectory inference for single cell RNA sequencing (scRNAseq) data is a powerful approach to understand time-dependent cellular processes such as cell cycle and cellular development. However, it is still not easy to infer the trajectory precisely by which cells differentiate to multiple lineages or exhibit cyclic transitions. Recent development of RNA velocity provides a way to visualize cell state transition without a prior knowledge. Trajectory inference that utilizes the velocity information will be highly useful to understand cellular dynamics.ResultsWe developed VeTra, a tool to infer the trajectories from scRNAseq data. Uniquely, VeTra can perform grouping of cells that are in the same stream of trajectory. For this, VeTra searches for weakly connected components of the directed graph obtained from RNA velocity. Therefore, VeTra makes it easy to define groups of cells from the origin and to the end of a certain trajectory. VeTra has been tested to infer the streams of cells for pancreatic development, neural development in hippocampus and cell cycle. VeTra is a useful tool to perform pseudo-time analysis from the start to the end of each group.

show abstract

Comprehensive single cell mRNA profiling reveals a detailed roadmap for pancreatic endocrinogenesis

Cited by 182 publications

References 62 publications

Generalizing RNA velocity to transient cell states through dynamical modeling

Generalizing RNA velocity to transient cell states through dynamical modeling

Preprocessing choices affect RNA velocity results for droplet scRNA-seq data

VeTra: a tool for trajectory inference based on RNA velocity

Contact Info

Product

Resources

About