Integration of a multi-omics stem cell differentiation dataset using a dynamical model
Ontology highlight
ABSTRACT:
ORGANISM(S): Mus musculus
PROVIDER: GSE93301 | GEO | 2023/04/25
REPOSITORIES: GEO
Similar Datasets
Project description:In eukaryotes, mRNA abundance is often a poor proxy for protein abundance. Despite this, the majority of methods used to dissect function in mammalian biology involve measurements of mRNA, with the assumption that changes observed at the mRNA level are reflected at the protein level. The inability to predict protein abundance from mRNA abundance is a major limitation when dissecting the relationship between genotype and phenotype in mammals. To understand the effect of cell-type, splice-variant selection and translation rate on the relationship between mRNA and protein abundance, we performed RNA-sequencing and mass-spectrometry proteomics of primary human naïve CD4+ T helper (NTH) cells at six time points during differentiation into T-helper type 1 (TH1) cells. For the RNA- sequencing, NTH cells were isolated from Human blood donors, subjected to T-helper 1 stimulation (anti-CD3/CD28, IL2 and IL12) for 0.5, 1, 2, 6 and 24 hours. At each time point cells were harvested, snap frozen and stored at -80 degrees Celsius. RNA was isolated from each sample and sequenced on an illumina 2500 instrument. By combining the information from the RNA- sequencing and the proteomics we constructed a simple mRNA-protein model, in which protein expression was defined as a linear combination of the splice variants of a gene, with a time-delay accounting for the dynamical effect induced by post-transcriptional processes and protein synthesis. This simple dynamical model resulted in a gene-protein correlation of rhoTH1 = 0.86, significantly higher than previously reported gene-protein prediction models in mammals.
2021-09-09 | E-MTAB-7775 | biostudies-arrayexpress
Project description:Mangrove-derived Streptomyces xiamenensis 318, with a relatively compact genome and simpler secondary metabolism, is used as model organism in our investigation. We performed integrated studies of metabolic dynamical modeling, transcriptome level measurements, and metabolic profiling experiments on this strain. To explore the relationship between primary and secondary metabolism, the global gene expression levels of strain 318 from early stationary phase to late stationary phase were compared by RNAseq analysis at 16 hour, 24 hour, 36 hour and 72 hour after batch culture started.
2019-02-10 | E-MTAB-7314 | biostudies-arrayexpress
Project description:Transcription factors (TFs) can relay signals through temporal alterations in their levels. The cell stress responsive TF p53 exhibits different dynamics depending on the type of stress, which influence the magnitude and dynamics of expression of its target genes and subsequent cellular outcomes. The mechanisms decoding p53 dynamics into levels and dynamics of RNA and protein of its targets remain unclear. We systematically quantified p53 target mRNA and protein levels over time under two p53 dynamical regimes – oscillatory and sustained – using RNA-seq and quantitative mass spectrometry. We categorized the mRNA dynamical patterns arising from oscillatory or sustained p53 expression, as well as the corresponding protein dynamics. We found that in both cases oscillatory dynamics allowed for the greatest variety of dynamical patterns of the downstream species. Target proteins from a wide range of functional categories were induced under both p53 dynamic conditions, with induction levels being overall higher under sustained dynamics. Mathematical modeling combined with analysis of empirical data revealed three mechanisms of decoding p53 dynamics: adjustment of mRNA and protein degradation rates, adjustment of activation thresholds for expression of mRNA and corresponding protein levels, and usage of coherent and incoherent feed-forward loop motifs in generating exclusive responses to p53 oscillatory or sustained dynamics, respectively. Our results highlight the diversity of mechanisms that decode complex TF dynamics into dynamical patterns of mRNA and proteins.
2022-02-11 | PXD027030 | Pride
Project description:Developmental cell fate decisions are dynamic processes driven by the complex behaviour of gene regulatory networks. A challenge in studying these processes using single-cell genomics is that the data provides only a static snapshot with no detail of dynamics. Metabolic labelling and splicing can provide time-resolved information, but current methods have limitations. Here, we present experimental and computational methods that overcome these limitations to allow dynamical modelling of gene expression from single-cell data. We developed sci-FATE2, an optimised metabolic labelling method that substantially increases data quality, and profiled approximately 45,000 embryonic stem cells differentiating into multiple neural tube identities. To recover dynamics, we developed velvet, a deep learning framework that extends beyond instantaneous velocity estimation by modelling gene expression dynamics through a neural stochastic differential equation system within a variational autoencoder. Velvet outperforms current velocity tools across quantitative benchmarks, and predicts trajectory distributions that accurately recapitulate underlying dataset distributions while conserving known biology. Velvet trajectory distributions capture dynamical aspects such as decision boundaries between alternative fates and correlative gene regulatory structure. Using velvet to provide a dynamical description of in vitro neural patterning, we highlight a process of sequential decision making and fate-specific patterns of developmental signalling. Together, these experimental and computational methods recast single-cell analyses from descriptions of observed data distributions to models of the dynamics that generated them, providing a new framework for investigating developmental gene regulation and cell fate decisions.
2023-07-06 | GSE236378 | GEO
Project description:Developmental cell fate decisions are dynamic processes driven by the complex behaviour of gene regulatory networks. A challenge in studying these processes using single-cell genomics is that the data provides only a static snapshot with no detail of dynamics. Metabolic labelling and splicing can provide time-resolved information, but current methods have limitations. Here, we present experimental and computational methods that overcome these limitations to allow dynamical modelling of gene expression from single-cell data. We developed sci-FATE2, an optimised metabolic labelling method that substantially increases data quality, and profiled approximately 45,000 embryonic stem cells differentiating into multiple neural tube identities. To recover dynamics, we developed velvet, a deep learning framework that extends beyond instantaneous velocity estimation by modelling gene expression dynamics through a neural stochastic differential equation system within a variational autoencoder. Velvet outperforms current velocity tools across quantitative benchmarks, and predicts trajectory distributions that accurately recapitulate underlying dataset distributions while conserving known biology. Velvet trajectory distributions capture dynamical aspects such as decision boundaries between alternative fates and correlative gene regulatory structure. Using velvet to provide a dynamical description of in vitro neural patterning, we highlight a process of sequential decision making and fate-specific patterns of developmental signalling. Together, these experimental and computational methods recast single-cell analyses from descriptions of observed data distributions to models of the dynamics that generated them, providing a new framework for investigating developmental gene regulation and cell fate decisions.
2023-07-06 | GSE236518 | GEO
Project description:Developmental cell fate decisions are dynamic processes driven by the complex behaviour of gene regulatory networks. A challenge in studying these processes using single-cell genomics is that the data provides only a static snapshot with no detail of dynamics. Metabolic labelling and splicing can provide time-resolved information, but current methods have limitations. Here, we present experimental and computational methods that overcome these limitations to allow dynamical modelling of gene expression from single-cell data. We developed sci-FATE2, an optimised metabolic labelling method that substantially increases data quality, and profiled approximately 45,000 embryonic stem cells differentiating into multiple neural tube identities. To recover dynamics, we developed velvet, a deep learning framework that extends beyond instantaneous velocity estimation by modelling gene expression dynamics through a neural stochastic differential equation system within a variational autoencoder. Velvet outperforms current velocity tools across quantitative benchmarks, and predicts trajectory distributions that accurately recapitulate underlying dataset distributions while conserving known biology. Velvet trajectory distributions capture dynamical aspects such as decision boundaries between alternative fates and correlative gene regulatory structure. Using velvet to provide a dynamical description of in vitro neural patterning, we highlight a process of sequential decision making and fate-specific patterns of developmental signalling. Together, these experimental and computational methods recast single-cell analyses from descriptions of observed data distributions to models of the dynamics that generated them, providing a new framework for investigating developmental gene regulation and cell fate decisions.
2023-07-06 | GSE236517 | GEO
Project description:Developmental cell fate decisions are dynamic processes driven by the complex behaviour of gene regulatory networks. A challenge in studying these processes using single-cell genomics is that the data provides only a static snapshot with no detail of dynamics. Metabolic labelling and splicing can provide time-resolved information, but current methods have limitations. Here, we present experimental and computational methods that overcome these limitations to allow dynamical modelling of gene expression from single-cell data. We developed sci-FATE2, an optimised metabolic labelling method that substantially increases data quality, and profiled approximately 45,000 embryonic stem cells differentiating into multiple neural tube identities. To recover dynamics, we developed velvet, a deep learning framework that extends beyond instantaneous velocity estimation by modelling gene expression dynamics through a neural stochastic differential equation system within a variational autoencoder. Velvet outperforms current velocity tools across quantitative benchmarks, and predicts trajectory distributions that accurately recapitulate underlying dataset distributions while conserving known biology. Velvet trajectory distributions capture dynamical aspects such as decision boundaries between alternative fates and correlative gene regulatory structure. Using velvet to provide a dynamical description of in vitro neural patterning, we highlight a process of sequential decision making and fate-specific patterns of developmental signalling. Together, these experimental and computational methods recast single-cell analyses from descriptions of observed data distributions to models of the dynamics that generated them, providing a new framework for investigating developmental gene regulation and cell fate decisions.
2023-07-06 | GSE236508 | GEO
Project description:Developmental cell fate decisions are dynamic processes driven by the complex behaviour of gene regulatory networks. A challenge in studying these processes using single-cell genomics is that the data provides only a static snapshot with no detail of dynamics. Metabolic labelling and splicing can provide time-resolved information, but current methods have limitations. Here, we present experimental and computational methods that overcome these limitations to allow dynamical modelling of gene expression from single-cell data. We developed sci-FATE2, an optimised metabolic labelling method that substantially increases data quality, and profiled approximately 45,000 embryonic stem cells differentiating into multiple neural tube identities. To recover dynamics, we developed velvet, a deep learning framework that extends beyond instantaneous velocity estimation by modelling gene expression dynamics through a neural stochastic differential equation system within a variational autoencoder. Velvet outperforms current velocity tools across quantitative benchmarks, and predicts trajectory distributions that accurately recapitulate underlying dataset distributions while conserving known biology. Velvet trajectory distributions capture dynamical aspects such as decision boundaries between alternative fates and correlative gene regulatory structure. Using velvet to provide a dynamical description of in vitro neural patterning, we highlight a process of sequential decision making and fate-specific patterns of developmental signalling. Together, these experimental and computational methods recast single-cell analyses from descriptions of observed data distributions to models of the dynamics that generated them, providing a new framework for investigating developmental gene regulation and cell fate decisions.
2023-07-06 | GSE236512 | GEO
Project description:Deep dynamical modelling of developmental trajectories with temporal transcriptomics
| PRJNA991515 | ENA
Project description:Differentiation of neural progenitor cells into mature neuronal phenotypes relies on extensive temporospatial coordination of mRNA expression to support the development of functional brain circuitry. Cleavage and polyadenylation of mRNA has tremendous regulatory capacity through the alteration of mRNA stability and modulation of microRNA (miRNA) function, however the extent of utilization in neuronal development is currently unclear. Here, we employed poly(A) tail sequencing, mRNA sequencing, ribosome profiling and small RNA sequencing to explore the functional relationship between mRNA abundance, translation, poly(A) tail length, alternative polyadenylation (APA) and miRNA expression in an in vitro model of neuronal differentiation. Differential analysis revealed a strong bias towards poly(A) tail and 3´UTR lengthening during differentiation, both of which were associated with changes in mRNA abundance but not translation. Globally, changes in miRNA expression were predominantly associated with mRNA abundance and translation, however several miRNA-mRNA pairings with potential to regulate poly(A) tail length were identified. Furthermore, 3´UTR lengthening was observed to significantly increase the inclusion of non-conserved miRNA binding sites, potentially enhancing the regulatory capacity of these molecules in mature neuronal cells. Together, our findings suggest poly(A) tail length and APA function as part of a rich post-transcriptional regulatory matrix during neuronal differentiation.
2023-06-09 | GSE155432 | GEO