Discovering governing equations of biological systems through representation learning and sparse model discovery

Understanding the governing rules of complex biological systems remains a significant challenge due to the nonlinear, high-dimensional nature of biological data. In this study, we present CLERA, a novel end-to-end computational framework designed to uncover parsimonious dynamical models and identify active gene programs from single-cell RNA sequencing data. By integrating a supervised autoencoder architecture with Sparse Identification of Nonlinear Dynamics, CLERA leverages prior knowledge to simultaneously extract related low-dimensional representation and uncover the underlying dynamical systems that drive the processes. Through the analysis of both synthetic and biological data, CLERA demonstrates robust performance in reconstructing gene expression dynamics, identifying key regulatory genes, and capturing temporal patterns across distinct cell types. CLERA's ability to generate dynamic interaction networks, combined with network rewiring using Personalized PageRank to highlight central genes and active gene programs, offers new insights into the complex regulatory mechanisms underlying cellular processes.

scCompass: An Integrated Multi-Species scRNA-seq Database for AI-Ready

Emerging single-cell sequencing technology has generated large amounts of data, allowing analysis of cellular dynamics and gene regulation at the single-cell resolution. Advances in artificial intelligence enhance life sciences research by delivering critical insights and optimizing data analysis processes. However, inconsistent data processing quality and standards remain to be a major challenge. Here scCompass is proposed, which provides a comprehensive resource designed to build large-scale, multi-species, and model-friendly single-cell data collection. By applying standardized data pre-processing, scCompass integrates and curates transcriptomic data from nearly 105 million single cells across 13 species. Using this extensive dataset, it is able to identify stable expression genes (SEGs) and organ-specific expression genes (OSGs) in humans and mice. Different scalable datasets are provided that can be easily adapted for AI model training and the pretrained checkpoints with state-of-the-art single-cell foundation models. In summary, scCompass is highly efficient and scalable database for AI-ready, which combined with user-friendly data sharing, visualization, and online analysis, greatly simplifies data access and exploitation for researchers in single-cell biology (http://www.bdbe.cn/kun).

Optimal transport reveals dynamic gene regulatory networks via gene velocity estimation

Inferring gene regulatory networks from gene expression data is an important and challenging problem in the biology community. We propose OTVelo, a methodology that takes time-stamped single-cell gene expression data as input and predicts gene regulation across two time points. It is known that the rate of change of gene expression, which we will refer to as gene velocity, provides crucial information that enhances such inference; however, this information is not always available due to the limitations in sequencing depth. Our algorithm overcomes this limitation by estimating gene velocities using optimal transport. We then infer gene regulation using time-lagged correlation and Granger causality via regularized linear regression. Instead of providing an aggregated network across all time points, our method uncovers the underlying dynamical mechanism across time points. We validate our algorithm on 13 simulated datasets with both synthetic and curated networks and demonstrate its efficacy on 9 experimental data sets.

Cell2fate infers RNA velocity modules to improve cell fate prediction

RNA velocity exploits the temporal information contained in spliced and unspliced RNA counts to infer transcriptional dynamics. Existing velocity models often rely on coarse biophysical simplifications or numerical approximations to solve the underlying ordinary differential equations (ODEs), which can compromise accuracy in challenging settings, such as complex or weak transcription rate changes across cellular trajectories. Here we present cell2fate, a formulation of RNA velocity based on a linearization of the velocity ODE, which allows solving a biophysically more accurate model in a fully Bayesian fashion. As a result, cell2fate decomposes the RNA velocity solutions into modules, providing a biophysical connection between RNA velocity and statistical dimensionality reduction. We comprehensively benchmark cell2fate in real-world settings, demonstrating enhanced interpretability and power to reconstruct complex dynamics and weak dynamical signals in rare and mature cell types. Finally, we apply cell2fate to the developing human brain, where we spatially map RNA velocity modules onto the tissue architecture, connecting the spatial organization of tissues with temporal dynamics of transcription.

The expression order determines the pioneer functions of NGN3 and NEUROD1 in pancreatic endocrine differentiation

Pioneer transcription factors (TFs) initiate chromatin remodeling, which is crucial for gene regulation and cell differentiation. In this study, we investigated how the sequential expression of neurogenin 3 (NGN3) and NEUROD1 affects their pioneering functions during pancreatic endocrine differentiation. Using a genetically engineered mouse model, we mapped NGN3-binding sites, confirming the pivotal role of this molecule in regulating chromatin accessibility. The pioneering function of NGN3 involves dose tolerance, and low doses are sufficient. Although NEUROD1 generally acts as a conventional TF, it can assume a pioneering role in the absence of NGN3. The sequential expression of NeuroD1 and Ngn3 predominantly drives α cell generation, which may explain the inefficient β cell induction observed in vitro. Our findings demonstrate that pioneer activity is dynamically shaped by temporal TF expression and inter-TF interactions, providing insights into transcriptional regulation and its implications for disease mechanisms and therapeutic targeting and enhancing in vitro differentiation strategies.

Identification of mouse and human embryonic pancreatic cells with adult Procr+ progenitor transcriptomic and epigenomic characteristics

Background

The quest to find a progenitor cell in the adult pancreas has driven research in the field for decades. Many potential progenitor cell sources have been reported, but so far this is a matter of debate mainly due to reproducibility issues. The existence of adult Procr+ progenitor cells in mice islets has been recently reported. These were shown to comprise ~1% of islet cells, lack expression of Neurog3 and endocrine hormones, and to be capable of differentiating into all endocrine cell types. However, these findings had limited impact, as further evidence supporting the existence and function of Procr+ progenitors has not emerged.

Methods and findings

We report here an unbiased comparison across mouse and human pancreatic samples, including adult islets and embryonic tissue, to track the existence of Procr+ progenitors originally described based on their global gene expression signature. We could not find Procr+ progenitors on other mouse or human adult pancreatic islet samples. Unexpectedly, our results revealed a transcriptionally close mesothelial cell population in the mouse and human embryonic pancreas. These Procr-like mesothelial cells of the embryonic pancreas share the salient transcriptional and epigenomic features of previously reported Procr+ progenitors found in adult pancreatic islets. Notably, we report here that Procr-like transcriptional signature is gradually established in mesothelial cells during mouse pancreas development from E12.5 to E17.5, which has its largest amount. Further supporting a developmentally relevant role in the human pancreas, we additionally report that a transcriptionally similar population is spontaneously differentiated from human pluripotent stem cells cultured in vitro along the pancreatic lineage.

Conclusions

Our results show that, although the previously reported Procr+ progenitor cell population could not be found in other adult pancreatic islet samples, a mesothelial cell population with a closely related transcriptional signature is present in both the mouse and human embryonic pancreas. Several lines of evidence presented in this work support a developmentally relevant function for these Procr-like mesothelial cells.

Mapping cells through time and space with moscot

Single-cell genomic technologies enable the multimodal profiling of millions of cells across temporal and spatial dimensions. However, experimental limitations hinder the comprehensive measurement of cells under native temporal dynamics and in their native spatial tissue niche. Optimal transport has emerged as a powerful tool to address these constraints and has facilitated the recovery of the original cellular context1-4. Yet, most optimal transport applications are unable to incorporate multimodal information or scale to single-cell atlases. Here we introduce multi-omics single-cell optimal transport (moscot), a scalable framework for optimal transport in single-cell genomics that supports multimodality across all applications. We demonstrate the capability of moscot to efficiently reconstruct developmental trajectories of 1.7 million cells from mouse embryos across 20 time points. To illustrate the capability of moscot in space, we enrich spatial transcriptomic datasets by mapping multimodal information from single-cell profiles in a mouse liver sample and align multiple coronal sections of the mouse brain. We present moscot.spatiotemporal, an approach that leverages gene-expression data across both spatial and temporal dimensions to uncover the spatiotemporal dynamics of mouse embryogenesis. We also resolve endocrine-lineage relationships of delta and epsilon cells in a previously unpublished mouse, time-resolved pancreas development dataset using paired measurements of gene expression and chromatin accessibility. Our findings are confirmed through experimental validation of NEUROD2 as a regulator of epsilon progenitor cells in a model of human induced pluripotent stem cell islet cell differentiation. Moscot is available as open-source software, accompanied by extensive documentation.

Diffusive topology preserving manifold distances for single-cell data analysis

Manifold learning techniques have emerged as crucial tools for uncovering latent patterns in high-dimensional single-cell data. However, most existing dimensionality reduction methods primarily rely on 2D visualization, which can distort true data relationships and fail to extract reliable biological information. Here, we present DTNE (diffusive topology neighbor embedding), a dimensionality reduction framework that faithfully approximates manifold distance to enhance cellular relationships and dynamics. DTNE constructs a manifold distance matrix using a modified personalized PageRank algorithm, thereby preserving topological structure while enabling diverse single-cell analyses. This approach facilitates distribution-based cellular relationship analysis, pseudotime inference, and clustering within a unified framework. Extensive benchmarking against mainstream algorithms on diverse datasets demonstrates DTNE's superior performance in maintaining geodesic distances and revealing significant biological patterns. Our results establish DTNE as a powerful tool for high-dimensional data analysis in uncovering meaningful biological insights.

Consensus representation of multiple cell-cell graphs from gene signaling pathways for cell type annotation

BACKGROUND

Recent advancements in single-cell RNA sequencing have greatly expanded our knowledge of the heterogeneous nature of tissues. However, robust and accurate cell type annotation continues to be a major challenge, hindered by issues such as marker specificity, batch effects, and a lack of comprehensive spatial and interaction data. Traditional annotation methods often fail to adequately address the complexity of cellular interactions and gene regulatory networks.

RESULTS

We proposed scMCGraph, a comprehensive computational framework that integrates gene expression with pathway activity to accurately annotate cell types within diverse scRNA-seq datasets. Initially, our model constructs multiple pathway-specific views using various pathway databases, which reflect both gene expression and pathway activities. These pathway-specific views are then integrated into a consensus graph. The consensus graph is subsequently utilized to reconstruct the multiple pathway views. Our model demonstrated exceptional robustness and accuracy across various analyses, including cross-platform, cross-time, cross-sample, and clinical dataset evaluations.

CONCLUSIONS

scMCGraph represents a significant advance in cell type annotation. The experiments have demonstrated that introducing pathway information significantly improves the learning of cell-cell graphs, with their resulting consensus graph enhancing the predictive performance of cell type prediction. Different pathway databases provide complementary data, and an increase in the number of pathways can also boost model performance. Extensive testing shows that in various cross-dataset application scenarios, scMCGraph consistently exhibits both accuracy and robustness.

Gene-level alignment of single-cell trajectories

Single-cell data analysis can infer dynamic changes in cell populations, for example across time, space or in response to perturbation, thus deriving pseudotime trajectories. Current approaches comparing trajectories often use dynamic programming but are limited by assumptions such as the existence of a definitive match. Here we describe Genes2Genes, a Bayesian information-theoretic dynamic programming framework for aligning single-cell trajectories. It is able to capture sequential matches and mismatches of individual genes between a reference and query trajectory, highlighting distinct clusters of alignment patterns. Across both real world and simulated datasets, it accurately inferred alignments and demonstrated its utility in disease cell-state trajectory analysis. In a proof-of-concept application, Genes2Genes revealed that T cells differentiated in vitro match an immature in vivo state while lacking expression of genes associated with TNF signaling. This demonstrates that precise trajectory alignment can pinpoint divergence from the in vivo system, thus guiding the optimization of in vitro culture conditions.

A statistical approach for systematic identification of transition cells from scRNA-seq data

Decoding cellular state transitions is crucial for understanding complex biological processes in development and disease. While recent advancements in single-cell RNA sequencing (scRNA-seq) offer insights into cellular trajectories, existing tools primarily study expressional rather than regulatory state shifts. We present CellTran, a statistical approach utilizing paired-gene expression correlations to detect transition cells from scRNA-seq data without explicitly resolving gene regulatory networks. Applying our approach to various contexts, including tissue regeneration, embryonic development, preinvasive lesions, and humoral responses post-vaccination, reveals transition cells and their distinct gene expression profiles. Our study sheds light on the underlying molecular mechanisms driving cellular state transitions, enhancing our ability to identify therapeutic targets for disease interventions.

Statistical inference with a manifold-constrained RNA velocity model uncovers cell cycle speed modulations

Across biological systems, cells undergo coordinated changes in gene expression, resulting in transcriptome dynamics that unfold within a low-dimensional manifold. While low-dimensional dynamics can be extracted using RNA velocity, these algorithms can be fragile and rely on heuristics lacking statistical control. Moreover, the estimated vector field is not dynamically consistent with the traversed gene expression manifold. To address these challenges, we introduce a Bayesian model of RNA velocity that couples velocity field and manifold estimation in a reformulated, unified framework, identifying the parameters of an explicit dynamical system. Focusing on the cell cycle, we implement VeloCycle to study gene regulation dynamics on one-dimensional periodic manifolds and validate its ability to infer cell cycle periods using live imaging. We also apply VeloCycle to reveal speed differences in regionally defined progenitors and Perturb-seq gene knockdowns. Overall, VeloCycle expands the single-cell RNA sequencing analysis toolkit with a modular and statistically consistent RNA velocity inference framework.

Population-Level Cell Trajectory Inference Based on Gaussian Distributions

In the past decade, inferring developmental trajectories from single-cell data has become a significant challenge in bioinformatics. RNA velocity, with its incorporation of directional dynamics, has significantly advanced the study of single-cell trajectories. However, as single-cell RNA sequencing technology evolves, it generates complex, high-dimensional data with high noise levels. Existing trajectory inference methods, which overlook cell distribution characteristics, may perform inadequately under such conditions. To address this, we introduce CPvGTI, a Gaussian distribution-based trajectory inference method. CPvGTI utilizes a Gaussian mixture model, optimized by the Expectation-Maximization algorithm, to construct new cell populations in the original data space. By integrating RNA velocity, CPvGTI employs Gaussian Process Regression to analyze the differentiation trajectories of these cell populations. To evaluate the performance of CPvGTI, we assess CPvGTI's performance against several state-of-the-art methods using four structurally diverse simulated datasets and four real datasets. The simulation studies indicate that CPvGTI excels in pseudo-time prediction and structural reconstruction compared to existing methods. Furthermore, the discovery of new branch trajectories in human forebrain and mouse hematopoiesis datasets confirms CPvGTI's superior performance.

Scalable identification of lineage-specific gene regulatory networks from metacells with NetID

The identification of gene regulatory networks (GRNs) is crucial for understanding cellular differentiation. Single-cell RNA sequencing data encode gene-level covariations at high resolution, yet data sparsity and high dimensionality hamper accurate and scalable GRN reconstruction. To overcome these challenges, we introduce NetID leveraging homogenous metacells while avoiding spurious gene-gene correlations. Benchmarking demonstrates superior performance of NetID compared to imputation-based methods. By incorporating cell fate probability information, NetID facilitates the prediction of lineage-specific GRNs and recovers known network motifs governing bone marrow hematopoiesis, making it a powerful toolkit for deciphering gene regulatory control of cellular differentiation from large-scale single-cell transcriptome data.

The SWI/SNF chromatin remodelling complex regulates pancreatic endocrine cell expansion and differentiation in mice in vivo

AIMS/HYPOTHESIS

Strategies to augment functional beta cell mass include directed differentiation of stem cells towards a beta cell fate, which requires extensive knowledge of transcriptional programs governing endocrine progenitor cell differentiation in vivo. We aimed to study the contributions of the Brahma-related gene-1 (BRG1) and Brahma (BRM) ATPase subunits of the SWI/SNF chromatin remodelling complex to endocrine cell development.

METHODS

We generated mice with endocrine progenitor-specific Neurog3-Cre BRG1 removal in the presence of heterozygous (Brg1Δendo;Brm+/-) or homozygous (double knockout: DKOΔendo) BRM deficiency. Whole-body metabolic phenotyping, islet function characterisation, islet quantitative PCR and histological characterisation were performed on animals and tissues postnatally. To test the mechanistic actions of SWI/SNF in controlling gene expression during endocrine cell development, single-cell RNA-seq was performed on flow-sorted endocrine-committed cells from embryonic day 15.5 control and mutant embryos.

RESULTS

Brg1Δendo;Brm+/- mice exhibit severe glucose intolerance, hyperglycaemia and hypoinsulinaemia, resulting, in part, from reduced islet number; diminished alpha, beta and delta cell mass; compromised islet insulin secretion; and altered islet gene expression programs, including reductions in MAFA and urocortin 3 (UCN3). DKOΔendo mice were not recovered at weaning; however, postnatal day 6 DKOΔendo mice were severely hyperglycaemic with reduced serum insulin levels and beta cell area. Single-cell RNA-seq of embryonic day 15.5 lineage-labelled cells revealed endocrine progenitor, alpha and beta cell populations from SWI/SNF mutants have reduced expression of Mafa, Gcg, Ins1 and Ins2, suggesting limited differentiation capacity. Reduced Neurog3 transcripts were discovered in DKOΔendo endocrine progenitor clusters, and the proliferative capacity of neurogenin 3 (NEUROG3)+ cells was reduced in Brg1Δendo;Brm+/- and DKOΔendo mutants.

CONCLUSIONS/INTERPRETATION

Loss of BRG1 from developing endocrine progenitor cells has a severe postnatal impact on glucose homeostasis, and loss of both subunits impedes animal survival, with both groups exhibiting alterations in hormone transcripts embryonically. Taken together, these data highlight the critical role SWI/SNF plays in governing gene expression programs essential for endocrine cell development and expansion.

DATA AVAILABILITY

Raw and processed data for scRNA-seq have been deposited into the NCBI Gene Expression Omnibus (GEO) database under the accession number GSE248369.

DeepKINET: a deep generative model for estimating single-cell RNA splicing and degradation rates

Messenger RNA splicing and degradation are critical for gene expression regulation, the abnormality of which leads to diseases. Previous methods for estimating kinetic rates have limitations, assuming uniform rates across cells. DeepKINET is a deep generative model that estimates splicing and degradation rates at single-cell resolution from scRNA-seq data. DeepKINET outperforms existing methods on simulated and metabolic labeling datasets. Applied to forebrain and breast cancer data, it identifies RNA-binding proteins responsible for kinetic rate diversity. DeepKINET also analyzes the effects of splicing factor mutations on target genes in erythroid lineage cells. DeepKINET effectively reveals cellular heterogeneity in post-transcriptional regulation.

FateNet: an integration of dynamical systems and deep learning for cell fate prediction

MOTIVATION

Understanding cellular decision-making, particularly its timing and impact on the biological system such as tissue health and function, is a fundamental challenge in biology and medicine. Existing methods for inferring fate decisions and cellular state dynamics from single-cell RNA sequencing data lack precision regarding decision points and broader tissue implications. Addressing this gap, we present FateNet, a computational approach integrating dynamical systems theory and deep learning to probe the cell decision-making process using scRNA-seq data.

RESULTS

By leveraging information about normal forms and scaling behavior near bifurcations common to many dynamical systems, FateNet predicts cell decision occurrence with higher accuracy than conventional methods and offers qualitative insights into the new state of the biological system. Also, through in-silico perturbation experiments, FateNet identifies key genes and pathways governing the differentiation process in hematopoiesis. Validated using different scRNA-seq data, FateNet emerges as a user-friendly and valuable tool for predicting critical points in biological processes, providing insights into complex trajectories.

AVAILABILITY AND IMPLEMENTATION

github.com/ThomasMBury/fatenet.

Variational inference of single cell time series

Time course single-cell RNA sequencing (scRNA-seq) enables researchers to probe genome-wide expression dynamics at the the single cell scale. However, when gene expression is affected jointly by time and cellular identity, analyzing such data - including conducting cell type annotation and modeling cell type-dependent dynamics - becomes challenging. To address this problem, we propose SNOW (SiNgle cell flOW map), a deep learning algorithm to deconvolve single cell time series data into time-dependent and time-independent contributions. SNOW has a number of advantages. First, it enables cell type annotation based on the time-independent dimensions. Second, it yields a probabilistic model that can be used to discriminate between biological temporal variation and batch effects contaminating individual timepoints, and provides an approach to mitigate batch effects. Finally, it is capable of projecting cells forward and backward in time, yielding time series at the individual cell level. This enables gene expression dynamics to be studied without the need for clustering or pseudobulking, which can be error prone and result in information loss. We describe our probabilistic framework in detail and demonstrate SNOW using data from three distinct time course scRNA-seq studies. Our results show that SNOW is able to construct biologically meaningful latent spaces, remove batch effects, and generate realistic time-series at the single-cell level. By way of example, we illustrate how the latter may be used to enhance the detection of cell type-specific circadian gene expression rhythms, and may be readily extended to other time-series analyses.

scPriorGraph: constructing biosemantic cell-cell graphs with prior gene set selection for cell type identification from scRNA-seq data

Cell type identification is an indispensable analytical step in single-cell data analyses. To address the high noise stemming from gene expression data, existing computational methods often overlook the biologically meaningful relationships between genes, opting to reduce all genes to a unified data space. We assume that such relationships can aid in characterizing cell type features and improving cell type recognition accuracy. To this end, we introduce scPriorGraph, a dual-channel graph neural network that integrates multi-level gene biosemantics. Experimental results demonstrate that scPriorGraph effectively aggregates feature values of similar cells using high-quality graphs, achieving state-of-the-art performance in cell type identification.

OmicVerse: a framework for bridging and deepening insights across bulk and single-cell sequencing

Single-cell sequencing is frequently affected by "omission" due to limitations in sequencing throughput, yet bulk RNA-seq may contain these ostensibly "omitted" cells. Here, we introduce the single cell trajectory blending from Bulk RNA-seq (BulkTrajBlend) algorithm, a component of the OmicVerse suite that leverages a Beta-Variational AutoEncoder for data deconvolution and graph neural networks for the discovery of overlapping communities. This approach effectively interpolates and restores the continuity of "omitted" cells within single-cell RNA sequencing datasets. Furthermore, OmicVerse provides an extensive toolkit for both bulk and single cell RNA-seq analysis, offering seamless access to diverse methodologies, streamlining computational processes, fostering exquisite data visualization, and facilitating the extraction of significant biological insights to advance scientific research.

Fatecode enables cell fate regulator prediction using classification-supervised autoencoder perturbation

Cell reprogramming, which guides the conversion between cell states, is a promising technology for tissue repair and regeneration, with the ultimate goal of accelerating recovery from diseases or injuries. To accomplish this, regulators must be identified and manipulated to control cell fate. We propose Fatecode, a computational method that predicts cell fate regulators based only on single-cell RNA sequencing (scRNA-seq) data. Fatecode learns a latent representation of the scRNA-seq data using a deep learning-based classification-supervised autoencoder and then performs in silico perturbation experiments on the latent representation to predict genes that, when perturbed, would alter the original cell type distribution to increase or decrease the population size of a cell type of interest. We assessed Fatecode's performance using simulations from a mechanistic gene-regulatory network model and scRNA-seq data mapping blood and brain development of different organisms. Our results suggest that Fatecode can detect known cell fate regulators from single-cell transcriptomics datasets.

Quantifying cell-state densities in single-cell phenotypic landscapes using Mellon

Cell-state density characterizes the distribution of cells along phenotypic landscapes and is crucial for unraveling the mechanisms that drive diverse biological processes. Here, we present Mellon, an algorithm for estimation of cell-state densities from high-dimensional representations of single-cell data. We demonstrate Mellon's efficacy by dissecting the density landscape of differentiating systems, revealing a consistent pattern of high-density regions corresponding to major cell types intertwined with low-density, rare transitory states. We present evidence implicating enhancer priming and the activation of master regulators in emergence of these transitory states. Mellon offers the flexibility to perform temporal interpolation of time-series data, providing a detailed view of cell-state dynamics during developmental processes. Mellon facilitates density estimation across various single-cell data modalities, scaling linearly with the number of cells. Our work underscores the importance of cell-state density in understanding the differentiation processes, and the potential of Mellon to provide insights into mechanisms guiding biological trajectories.

Optimal sequencing budget allocation for trajectory reconstruction of single cells

BACKGROUND

Charting cellular trajectories over gene expression is key to understanding dynamic cellular processes and their underlying mechanisms. While advances in single-cell RNA-sequencing technologies and computational methods have pushed forward the recovery of such trajectories, trajectory inference remains a challenge due to the noisy, sparse, and high-dimensional nature of single-cell data. This challenge can be alleviated by increasing either the number of cells sampled along the trajectory (breadth) or the sequencing depth, i.e. the number of reads captured per cell (depth). Generally, these two factors are coupled due to an inherent breadth-depth tradeoff that arises when the sequencing budget is constrained due to financial or technical limitations.

RESULTS

Here we study the optimal allocation of a fixed sequencing budget to optimize the recovery of trajectory attributes. Empirical results reveal that reconstruction accuracy of internal cell structure in expression space scales with the logarithm of either the breadth or depth of sequencing. We additionally observe a power law relationship between the optimal number of sampled cells and the corresponding sequencing budget. For linear trajectories, non-monotonicity in trajectory reconstruction across the breadth-depth tradeoff can impact downstream inference, such as expression pattern analysis along the trajectory. We demonstrate these results for five single-cell RNA-sequencing datasets encompassing differentiation of embryonic stem cells, pancreatic beta cells, hepatoblast and multipotent hematopoietic cells, as well as induced reprogramming of embryonic fibroblasts into neurons. By addressing the challenges of single-cell data, our study offers insights into maximizing the efficiency of cellular trajectory analysis through strategic allocation of sequencing resources.

Dissection and integration of bursty transcriptional dynamics for complex systems

RNA velocity estimation is a potentially powerful tool to reveal the directionality of transcriptional changes in single-cell RNA-sequencing data, but it lacks accuracy, absent advanced metabolic labeling techniques. We developed an approach, TopicVelo, that disentangles simultaneous, yet distinct, dynamics by using a probabilistic topic model, a highly interpretable form of latent space factorization, to infer cells and genes associated with individual processes, thereby capturing cellular pluripotency or multifaceted functionality. Focusing on process-associated cells and genes enables accurate estimation of process-specific velocities via a master equation for a transcriptional burst model accounting for intrinsic stochasticity. The method obtains a global transition matrix by leveraging cell topic weights to integrate process-specific signals. In challenging systems, this method accurately recovers complex transitions and terminal states, while our use of first-passage time analysis provides insights into transient transitions. These results expand the limits of RNA velocity, empowering future studies of cell fate and functional responses.

Modeling gene expression cascades during cell state transitions

During cellular processes such as differentiation or response to external stimuli, cells exhibit dynamic changes in their gene expression profiles. Single-cell RNA sequencing (scRNA-seq) can be used to investigate these dynamic changes. To this end, cells are typically ordered along a pseudotemporal trajectory which recapitulates the progression of cells as they transition from one cell state to another. We infer transcriptional dynamics by modeling the gene expression profiles in pseudotemporally ordered cells using a Bayesian inference approach. This enables ordering genes along transcriptional cascades, estimating differences in the timing of gene expression dynamics, and deducing regulatory gene interactions. Here, we apply this approach to scRNA-seq datasets derived from mouse embryonic forebrain and pancreas samples. This analysis demonstrates the utility of the method to derive the ordering of gene dynamics and regulatory relationships critical for proper cellular differentiation and maturation across a variety of developmental contexts.

A deep learning adversarial autoencoder with dynamic batching displays high performance in denoising and ordering scRNA-seq data

By providing high-resolution of cell-to-cell variation in gene expression, single-cell RNA sequencing (scRNA-seq) offers insights into cell heterogeneity, differentiating dynamics, and disease mechanisms. However, challenges such as low capture rates and dropout events can introduce noise in data analysis. Here, we propose a deep neural generative framework, the dynamic batching adversarial autoencoder (DB-AAE), which excels at denoising scRNA-seq datasets. DB-AAE directly captures optimal features from input data and enhances feature preservation, including cell type-specific gene expression patterns. Comprehensive evaluation on simulated and real datasets demonstrates that DB-AAE outperforms other methods in denoising accuracy and biological signal preservation. It also improves the accuracy of other algorithms in establishing pseudo-time inference. This study highlights DB-AAE's effectiveness and potential as a valuable tool for enhancing the quality and reliability of downstream analyses in scRNA-seq research.

TFvelo: gene regulation inspired RNA velocity estimation

RNA velocity is closely related with cell fate and is an important indicator for the prediction of cell states with elegant physical explanation derived from single-cell RNA-seq data. Most existing RNA velocity models aim to extract dynamics from the phase delay between unspliced and spliced mRNA for each individual gene. However, unspliced/spliced mRNA abundance may not provide sufficient signal for dynamic modeling, leading to poor fit in phase portraits. Motivated by the idea that RNA velocity could be driven by the transcriptional regulation, we propose TFvelo, which expands RNA velocity concept to various single-cell datasets without relying on splicing information, by introducing gene regulatory information. Our experiments on synthetic data and multiple scRNA-Seq datasets show that TFvelo can accurately fit genes dynamics on phase portraits, and effectively infer cell pseudo-time and trajectory from RNA abundance data. TFvelo opens a robust and accurate avenue for modeling RNA velocity for single cell data.

Statistical inference with a manifold-constrained RNA velocity model uncovers cell cycle speed modulations

Across a range of biological processes, cells undergo coordinated changes in gene expression, resulting in transcriptome dynamics that unfold within a low-dimensional manifold. Single-cell RNA-sequencing (scRNA-seq) only measures temporal snapshots of gene expression. However, information on the underlying low-dimensional dynamics can be extracted using RNA velocity, which models unspliced and spliced RNA abundances to estimate the rate of change of gene expression. Available RNA velocity algorithms can be fragile and rely on heuristics that lack statistical control. Moreover, the estimated vector field is not dynamically consistent with the traversed gene expression manifold. Here, we develop a generative model of RNA velocity and a Bayesian inference approach that solves these problems. Our model couples velocity field and manifold estimation in a reformulated, unified framework, so as to coherently identify the parameters of an autonomous dynamical system. Focusing on the cell cycle, we implemented VeloCycle to study gene regulation dynamics on one-dimensional periodic manifolds and validated using live-imaging its ability to infer actual cell cycle periods. We benchmarked RNA velocity inference with sensitivity analyses and demonstrated one- and multiple-sample testing. We also conducted Markov chain Monte Carlo inference on the model, uncovering key relationships between gene-specific kinetics and our gene-independent velocity estimate. Finally, we applied VeloCycle to in vivo samples and in vitro genome-wide Perturb-seq, revealing regionally-defined proliferation modes in neural progenitors and the effect of gene knockdowns on cell cycle speed. Ultimately, VeloCycle expands the scRNA-seq analysis toolkit with a modular and statistically rigorous RNA velocity inference framework.

DeepVelo: deep learning extends RNA velocity to multi-lineage systems with cell-specific kinetics

Existing RNA velocity estimation methods strongly rely on predefined dynamics and cell-agnostic constant transcriptional kinetic rates, assumptions often violated in complex and heterogeneous single-cell RNA sequencing (scRNA-seq) data. Using a graph convolution network, DeepVelo overcomes these limitations by generalizing RNA velocity to cell populations containing time-dependent kinetics and multiple lineages. DeepVelo infers time-varying cellular rates of transcription, splicing, and degradation, recovers each cell's stage in the differentiation process, and detects functionally relevant driver genes regulating these processes. Application to various developmental and pathogenic processes demonstrates DeepVelo's capacity to study complex differentiation and lineage decision events in heterogeneous scRNA-seq data.

A relay velocity model infers cell-dependent RNA velocity

RNA velocity provides an approach for inferring cellular state transitions from single-cell RNA sequencing (scRNA-seq) data. Conventional RNA velocity models infer universal kinetics from all cells in an scRNA-seq experiment, resulting in unpredictable performance in experiments with multi-stage and/or multi-lineage transition of cell states where the assumption of the same kinetic rates for all cells no longer holds. Here we present cellDancer, a scalable deep neural network that locally infers velocity for each cell from its neighbors and then relays a series of local velocities to provide single-cell resolution inference of velocity kinetics. In the simulation benchmark, cellDancer shows robust performance in multiple kinetic regimes, high dropout ratio datasets and sparse datasets. We show that cellDancer overcomes the limitations of existing RNA velocity models in modeling erythroid maturation and hippocampus development. Moreover, cellDancer provides cell-specific predictions of transcription, splicing and degradation rates, which we identify as potential indicators of cell fate in the mouse pancreas.

Deep generative modeling of transcriptional dynamics for RNA velocity analysis in single cells

RNA velocity has been rapidly adopted to guide interpretation of transcriptional dynamics in snapshot single-cell data; however, current approaches for estimating RNA velocity lack effective strategies for quantifying uncertainty and determining the overall applicability to the system of interest. Here, we present veloVI (velocity variational inference), a deep generative modeling framework for estimating RNA velocity. veloVI learns a gene-specific dynamical model of RNA metabolism and provides a transcriptome-wide quantification of velocity uncertainty. We show that veloVI compares favorably to previous approaches with respect to goodness of fit, consistency across transcriptionally similar cells and stability across preprocessing pipelines for quantifying RNA abundance. Further, we demonstrate that veloVI's posterior velocity uncertainty can be used to assess whether velocity analysis is appropriate for a given dataset. Finally, we highlight veloVI as a flexible framework for modeling transcriptional dynamics by adapting the underlying dynamical model to use time-dependent transcription rates.

Interpretable trajectory inference with single-cell Linear Adaptive Negative-binomial Expression (scLANE) testing

The rapid proliferation of trajectory inference methods for single-cell RNA-seq data has allowed researchers to investigate complex biological processes by examining underlying gene expression dynamics. After estimating a latent cell ordering, statistical models are used to determine which genes exhibit changes in expression that are significantly associated with progression through the biological trajectory. While a few techniques for performing trajectory differential expression exist, most rely on the flexibility of generalized additive models in order to account for the inherent nonlinearity of changes in gene expression. As such, the results can be difficult to interpret, and biological conclusions often rest on subjective visual inspections of the most dynamic genes. To address this challenge, we propose scLANE testing, which is built around an interpretable generalized linear model and handles nonlinearity with basis splines chosen empirically for each gene. In addition, extensions to estimating equations and mixed models allow for reliable trajectory testing under complex experimental designs. After validating the accuracy of scLANE under several different simulation scenarios, we apply it to a set of diverse biological datasets and display its ability to provide novel biological information when used downstream of both pseudotime and RNA velocity estimation methods.

SPIDE: A single cell potency inference method based on the local cell-specific network entropy

For a given single cell RNA-seq data, it is critical to pinpoint key cellular stages and quantify cells' differentiation potency along a differentiation pathway in a time course manner. Currently, several methods based on the entropy of gene functions or PPI network have been proposed to solve the problem. Nevertheless, these methods still suffer from the inaccurate interactions and noises originating from scRNA-seq profile. In this study, we proposed a cell potency inference method based on cell-specific network entropy, called SPIDE. SPIDE introduces the local weighted cell-specific network for each cell to maintain cell heterogeneity and calculates the entropy by incorporating gene expression with network structure. In this study, we compared three cell entropy estimation models on eight scRNA-Seq datasets. The results show that SPIDE obtains consistent conclusions with real cell differentiation potency on most datasets. Moreover, SPIDE accurately recovers the continuous changes of potency during cell differentiation and significantly correlates with the stemness of tumor cells in Colorectal cancer. To conclude, our study provides a universal and accurate framework for cell entropy estimation, which deepens our understanding of cell differentiation, the development of diseases and other related biological research.

Pumping the brakes on RNA velocity by understanding and interpreting RNA velocity estimates

BACKGROUND

RNA velocity analysis of single cells offers the potential to predict temporal dynamics from gene expression. In many systems, RNA velocity has been observed to produce a vector field that qualitatively reflects known features of the system. However, the limitations of RNA velocity estimates are still not well understood.

RESULTS

We analyze the impact of different steps in the RNA velocity workflow on direction and speed. We consider both high-dimensional velocity estimates and low-dimensional velocity vector fields mapped onto an embedding. We conclude the transition probability method for mapping velocity estimates onto an embedding is effectively interpolating in the embedding space. Our findings reveal a significant dependence of the RNA velocity workflow on smoothing via the k-nearest-neighbors (k-NN) graph of the observed data. This reliance results in considerable estimation errors for both direction and speed in both high- and low-dimensional settings when the k-NN graph fails to accurately represent the true data structure; this is an unknown feature of real data. RNA velocity performs poorly at estimating speed in both low- and high-dimensional spaces, except in very low noise settings. We introduce a novel quality measure that can identify when RNA velocity should not be used.

CONCLUSIONS

Our findings emphasize the importance of choices in the RNA velocity workflow and highlight critical limitations of data analysis. We advise against over-interpreting expression dynamics using RNA velocity, particularly in terms of speed. Finally, we emphasize that the use of RNA velocity in assessing the correctness of a low-dimensional embedding is circular.

NEUROD2 function is dispensable for human pancreatic β cell specification

Introduction

The molecular programs regulating human pancreatic endocrine cell induction and fate allocation are not well deciphered. Here, we investigated the spatiotemporal expression pattern and the function of the neurogenic differentiation factor 2 (NEUROD2) during human endocrinogenesis.

Methods

Using Crispr-Cas9 gene editing, we generated a reporter knock-in transcription factor (TF) knock-out human inducible pluripotent stem cell (iPSC) line in which the open reading frame of both NEUROD2 alleles are replaced by a nuclear histone 2B-Venus reporter (NEUROD2nVenus/nVenus).

Results

We identified a transient expression of NEUROD2 mRNA and its nuclear Venus reporter activity at the stage of human endocrine progenitor formation in an iPSC differentiation model. This expression profile is similar to what was previously reported in mice, uncovering an evolutionarily conserved gene expression pattern of NEUROD2 during endocrinogenesis. In vitro differentiation of the generated homozygous NEUROD2nVenus/nVenus iPSC line towards human endocrine lineages uncovered no significant impact upon the loss of NEUROD2 on endocrine cell induction. Moreover, analysis of endocrine cell specification revealed no striking changes in the generation of insulin-producing b cells and glucagon-secreting a cells upon lack of NEUROD2.

Discussion

Overall, our results suggest that NEUROD2 is expendable for human b cell formation in vitro.

Deciphering early human pancreas development at the single-cell level

Understanding pancreas development can provide clues for better treatments of pancreatic diseases. However, the molecular heterogeneity and developmental trajectory of the early human pancreas are poorly explored. Here, we performed large-scale single-cell RNA sequencing and single-cell assay for transposase accessible chromatin sequencing of human embryonic pancreas tissue obtained from first-trimester embryos. We unraveled the molecular heterogeneity, developmental trajectories and regulatory networks of the major cell types. The results reveal that dorsal pancreatic multipotent cells in humans exhibit different gene expression patterns than ventral multipotent cells. Pancreato-biliary progenitors that generate ventral multipotent cells in humans were identified. Notch and MAPK signals from mesenchymal cells regulate the differentiation of multipotent cells into trunk and duct cells. Notably, we identified endocrine progenitor subclusters with different differentiation potentials. Although the developmental trajectories are largely conserved between humans and mice, some distinct gene expression patterns have also been identified. Overall, we provide a comprehensive landscape of early human pancreas development to understand its lineage transitions and molecular complexity.

Delineating mouse β-cell identity during lifetime and in diabetes with a single cell atlas

Although multiple pancreatic islet single-cell RNA-sequencing (scRNA-seq) datasets have been generated, a consensus on pancreatic cell states in development, homeostasis and diabetes as well as the value of preclinical animal models is missing. Here, we present an scRNA-seq cross-condition mouse islet atlas (MIA), a curated resource for interactive exploration and computational querying. We integrate over 300,000 cells from nine scRNA-seq datasets consisting of 56 samples, varying in age, sex and diabetes models, including an autoimmune type 1 diabetes model (NOD), a glucotoxicity/lipotoxicity type 2 diabetes model (db/db) and a chemical streptozotocin β-cell ablation model. The β-cell landscape of MIA reveals new cell states during disease progression and cross-publication differences between previously suggested marker genes. We show that β-cells in the streptozotocin model transcriptionally correlate with those in human type 2 diabetes and mouse db/db models, but are less similar to human type 1 diabetes and mouse NOD β-cells. We also report pathways that are shared between β-cells in immature, aged and diabetes models. MIA enables a comprehensive analysis of β-cell responses to different stressors, providing a roadmap for the understanding of β-cell plasticity, compensation and demise.

Coordination between ECM and cell-cell adhesion regulates the development of islet aggregation, architecture, and functional maturation

Pancreatic islets are three-dimensional cell aggregates consisting of unique cellular composition, cell-to-cell contacts, and interactions with blood vessels. Cell aggregation is essential for islet endocrine function; however, it remains unclear how developing islets establish aggregation. By combining genetic animal models, imaging tools, and gene expression profiling, we demonstrate that islet aggregation is regulated by extracellular matrix signaling and cell-cell adhesion. Islet endocrine cell-specific inactivation of extracellular matrix receptor integrin β1 disrupted blood vessel interactions but promoted cell-cell adhesion and the formation of larger islets. In contrast, ablation of cell-cell adhesion molecule α-catenin promoted blood vessel interactions yet compromised islet clustering. Simultaneous removal of integrin β1 and α-catenin disrupts islet aggregation and the endocrine cell maturation process, demonstrating that establishment of islet aggregates is essential for functional maturation. Our study provides new insights into understanding the fundamental self-organizing mechanism for islet aggregation, architecture, and functional maturation.

Characterization of cell-fate decision landscapes by estimating transcription factor dynamics

Time-specific modulation of gene expression during differentiation by transcription factors promotes cell diversity. However, estimating their dynamic regulatory activity at the single-cell level and in a high-throughput manner remains challenging. We present FateCompass, an integrative approach that utilizes single-cell transcriptomics data to identify lineage-specific transcription factors throughout differentiation. By combining a probabilistic framework with RNA velocities or differentiation potential, we estimate transition probabilities, while a linear model of gene regulation is employed to compute transcription factor activities. Considering dynamic changes and correlations of expression and activities, FateCompass identifies lineage-specific regulators. Our validation using in silico data and application to pancreatic endocrine cell differentiation datasets highlight both known and potentially novel lineage-specific regulators. Notably, we uncovered undescribed transcription factors of an enterochromaffin-like population during in vitro differentiation toward ß-like cells. FateCompass provides a valuable framework for hypothesis generation, advancing our understanding of the gene regulatory networks driving cell-fate decisions.

Quantifying Cell-State Densities in Single-Cell Phenotypic Landscapes using Mellon

Cell-state density characterizes the distribution of cells along phenotypic landscapes and is crucial for unraveling the mechanisms that drive cellular differentiation, regeneration, and disease. Here, we present Mellon, a novel computational algorithm for high-resolution estimation of cell-state densities from single-cell data. We demonstrate Mellon's efficacy by dissecting the density landscape of various differentiating systems, revealing a consistent pattern of high-density regions corresponding to major cell types intertwined with low-density, rare transitory states. Utilizing hematopoietic stem cell fate specification to B-cells as a case study, we present evidence implicating enhancer priming and the activation of master regulators in the emergence of these transitory states. Mellon offers the flexibility to perform temporal interpolation of time-series data, providing a detailed view of cell-state dynamics during the inherently continuous developmental processes. Scalable and adaptable, Mellon facilitates density estimation across various single-cell data modalities, scaling linearly with the number of cells. Our work underscores the importance of cell-state density in understanding the differentiation processes, and the potential of Mellon to provide new insights into the regulatory mechanisms guiding cellular fate decisions.

scTour: a deep learning architecture for robust inference and accurate prediction of cellular dynamics

Despite the continued efforts, a batch-insensitive tool that can both infer and predict the developmental dynamics using single-cell genomics is lacking. Here, I present scTour, a novel deep learning architecture to perform robust inference and accurate prediction of cellular dynamics with minimal influence from batch effects. For inference, scTour simultaneously estimates the developmental pseudotime, delineates the vector field, and maps the transcriptomic latent space under a single, integrated framework. For prediction, scTour precisely reconstructs the underlying dynamics of unseen cellular states or a new independent dataset. scTour's functionalities are demonstrated in a variety of biological processes from 19 datasets.

ISL1 controls pancreatic alpha cell fate and beta cell maturation

BACKGROUND

Glucose homeostasis is dependent on functional pancreatic α and ß cells. The mechanisms underlying the generation and maturation of these endocrine cells remain unclear.

RESULTS

We unravel the molecular mode of action of ISL1 in controlling α cell fate and the formation of functional ß cells in the pancreas. By combining transgenic mouse models, transcriptomic and epigenomic profiling, we uncover that elimination of Isl1 results in a diabetic phenotype with a complete loss of α cells, disrupted pancreatic islet architecture, downregulation of key ß-cell regulators and maturation markers of ß cells, and an enrichment in an intermediate endocrine progenitor transcriptomic profile.

CONCLUSIONS

Mechanistically, apart from the altered transcriptome of pancreatic endocrine cells, Isl1 elimination results in altered silencing H3K27me3 histone modifications in the promoter regions of genes that are essential for endocrine cell differentiation. Our results thus show that ISL1 transcriptionally and epigenetically controls α cell fate competence, and ß cell maturation, suggesting that ISL1 is a critical component for generating functional α and ß cells.

Biologically informed deep learning to query gene programs in single-cell atlases

The increasing availability of large-scale single-cell atlases has enabled the detailed description of cell states. In parallel, advances in deep learning allow rapid analysis of newly generated query datasets by mapping them into reference atlases. However, existing data transformations learned to map query data are not easily explainable using biologically known concepts such as genes or pathways. Here we propose expiMap, a biologically informed deep-learning architecture that enables single-cell reference mapping. ExpiMap learns to map cells into biologically understandable components representing known 'gene programs'. The activity of each cell for a gene program is learned while simultaneously refining them and learning de novo programs. We show that expiMap compares favourably to existing methods while bringing an additional layer of interpretability to integrative single-cell analysis. Furthermore, we demonstrate its applicability to analyse single-cell perturbation responses in different tissues and species and resolve responses of patients who have coronavirus disease 2019 to different treatments across cell types.

Transformer for one stop interpretable cell type annotation

Consistent annotation transfer from reference dataset to query dataset is fundamental to the development and reproducibility of single-cell research. Compared with traditional annotation methods, deep learning based methods are faster and more automated. A series of useful single cell analysis tools based on autoencoder architecture have been developed but these struggle to strike a balance between depth and interpretability. Here, we present TOSICA, a multi-head self-attention deep learning model based on Transformer that enables interpretable cell type annotation using biologically understandable entities, such as pathways or regulons. We show that TOSICA achieves fast and accurate one-stop annotation and batch-insensitive integration while providing biologically interpretable insights for understanding cellular behavior during development and disease progressions. We demonstrate TOSICA's advantages by applying it to scRNA-seq data of tumor-infiltrating immune cells, and CD14+ monocytes in COVID-19 to reveal rare cell types, heterogeneity and dynamic trajectories associated with disease progression and severity.

Dynamic visualization of high-dimensional data

Dimensionality reduction (DR) is commonly used to project high-dimensional data into lower dimensions for visualization, which could then generate new insights and hypotheses. However, DR algorithms introduce distortions in the visualization and cannot faithfully represent all relations in the data. Thus, there is a need for methods to assess the reliability of DR visualizations. Here we present DynamicViz, a framework for generating dynamic visualizations that capture the sensitivity of DR visualizations to perturbations in the data resulting from bootstrap sampling. DynamicViz can be applied to all commonly used DR methods. We show the utility of dynamic visualizations in diagnosing common interpretative pitfalls of static visualizations and extending existing single-cell analyses. We introduce the variance score to quantify the dynamic variability of observations in these visualizations. The variance score characterizes natural variability in the data and can be used to optimize DR algorithm implementations.

A Guide to Trajectory Inference and RNA Velocity

Technological developments have led to an explosion of high-throughput single-cell data, which are revealing unprecedented perspectives on cell identity. Recently, significant attention has focused on investigating, from single-cell RNA-sequencing (scRNA-seq) data, cellular dynamic processes, such as cell differentiation, cell cycle and cell (de)activation. In particular, trajectory inference methods, by ordering cells along a trajectory, allow estimating a differentiation tree of cells. While trajectory inference tools typically work with gene expression levels, common scRNA-seq protocols allow the identification and quantification of unspliced pre-mRNAs and mature spliced mRNAs for each gene. By exploiting the abundance of unspliced and spliced mRNA, one can infer the RNA velocity of individual cells, i.e., the time derivative of the gene expression state of cells. Whereas traditional trajectory inference methods reconstruct cellular dynamics given a population of cells of varying maturity, RNA velocity relies on a dynamical model describing splicing dynamics. Here, we initially discuss conceptual and theoretical aspects of both approaches, then illustrate how they can be combined together, and finally present an example use case on real data.

DeepVelo: Single-cell transcriptomic deep velocity field learning with neural ordinary differential equations

Recent advances in single-cell sequencing technologies have provided unprecedented opportunities to measure the gene expression profile and RNA velocity of individual cells. However, modeling transcriptional dynamics is computationally challenging because of the high-dimensional, sparse nature of the single-cell gene expression measurements and the nonlinear regulatory relationships. Here, we present DeepVelo, a neural network-based ordinary differential equation that can model complex transcriptome dynamics by describing continuous-time gene expression changes within individual cells. We apply DeepVelo to public datasets from different sequencing platforms to (i) formulate transcriptome dynamics on different time scales, (ii) measure the instability of cell states, and (iii) identify developmental driver genes via perturbation analysis. Benchmarking against the state-of-the-art methods shows that DeepVelo can learn a more accurate representation of the velocity field. Furthermore, our perturbation studies reveal that single-cell dynamical systems could exhibit chaotic properties. In summary, DeepVelo allows data-driven discoveries of differential equations that delineate single-cell transcriptome dynamics.

Cascade Forest-Based Model for Prediction of RNA Velocity

In recent years, single-cell RNA sequencing technology (scRNA-seq) has developed rapidly and has been widely used in biological and medical research, such as in expression heterogeneity and transcriptome dynamics of single cells. The investigation of RNA velocity is a new topic in the study of cellular dynamics using single-cell RNA sequencing data. It can recover directional dynamic information from single-cell transcriptomics by linking measurements to the underlying dynamics of gene expression. Predicting the RNA velocity vector of each cell based on its gene expression data and formulating RNA velocity prediction as a classification problem is a new research direction. In this paper, we develop a cascade forest model to predict RNA velocity. Compared with other popular ensemble classifiers, such as XGBoost, RandomForest, LightGBM, NGBoost, and TabNet, it performs better in predicting RNA velocity. This paper provides guidance for researchers in selecting and applying appropriate classification tools in their analytical work and suggests some possible directions for future improvement of classification tools.

spliceJAC: transition genes and state-specific gene regulation from single-cell transcriptome data

Extracting dynamical information from single-cell transcriptomics is a novel task with the promise to advance our understanding of cell state transition and interactions between genes. Yet, theory-oriented, bottom-up approaches that consider differences among cell states are largely lacking. Here, we present spliceJAC, a method to quantify the multivariate mRNA splicing from single-cell RNA sequencing (scRNA-seq). spliceJAC utilizes the unspliced and spliced mRNA count matrices to constructs cell state-specific gene-gene regulatory interactions and applies stability analysis to predict putative driver genes critical to the transitions between cell states. By applying spliceJAC to biological systems including pancreas endothelium development and epithelial-mesenchymal transition (EMT) in A549 lung cancer cells, we predict genes that serve specific signaling roles in different cell states, recover important differentially expressed genes in agreement with pre-existing analysis, and predict new transition genes that are either exclusive or shared between different cell state transitions.

Towards reliable quantification of cell state velocities

A few years ago, it was proposed to use the simultaneous quantification of unspliced and spliced messenger RNA (mRNA) to add a temporal dimension to high-throughput snapshots of single cell RNA sequencing data. This concept can yield additional insight into the transcriptional dynamics of the biological systems under study. However, current methods for inferring cell state velocities from such data (known as RNA velocities) are afflicted by several theoretical and computational problems, hindering realistic and reliable velocity estimation. We discuss these issues and propose new solutions for addressing some of the current challenges in consistency of data processing, velocity inference and visualisation. We translate our computational conclusion in two velocity analysis tools: one detailed method κ-velo and one heuristic method eco-velo, each of which uses a different set of assumptions about the data.