GReCS · Characterizing gene regulation in single cells through integration of scRNA-seq and scATAC-seq data with generic multi-modal prior information
Horizon 2020 — Marie Skłodowska-Curie Actions
- Duration
- 2021-11-15 → 2023-11-14
- EU contribution
- €224,934
- Participants
- 1
- Scheme
- MSCA-IF
Lines connect the coordinator with its partners.
Results in brief
Characterizing gene regulation in single cells through integration of scRNA-seq and scATAC-seq data with generic multi-modal prior information
The overarching goal of this project was to work towards a better understanding of how gene regulation varies across different cell types and states. The advent of single cell technologies has enabled the characterisation of cell states at a more granular level. In particular, scRNA- and scATAC-seq data have become more widely used and large atlases of gene expression and chromatin accessibility at single cell resolution are being assembled. As part of this project single cell RNA and ATAC datasets were used to infer gene regulatory networks (GRNs), consisting of transcription factors (TFs) and their target genes, as well as enhancers harbouring TF binding sites. Using the data at single cell resolution, these enhancer-GRNs were further analysed for differences between cell states. As one part of the project, a reference dataset of transcription factor regulation had to be assembled using an extensive list of published datasets. A second part involved developing a method to identify relevant connections in a given network by integrating different modalities and cell type specific input data. Finally, single cell RNA and ATAC data were to be jointly used to infer gene regulatory networks in a new data analysis application. Overall, all objectives could be fulfilled, with the exception of minor changes due to new scientific developments since the start of the action.
Data: CORDIS, © European Union
Project objective
The advent of single cell technologies has enabled the characterization of cell types and developmental processes. Observations from different cells allow one to identify underlying patterns at higher resolution than convoluted bulk data, and integration of different omics data can yield a more differentiated picture of mechanistic connections. In this proposal, Gene REgulatory Cell States (GReCS) from multi-modal data, I plan to develop a computational method that combines these aspects to generate insights into gene regulation at the level of single cells.Measurements of chromatin accessibility in single cells are becoming increasingly common. The method I propose to develop combines sc/sn-ATAC- and scRNA-sequencing data to characterize gene regulation. My approach will integrate and use transcriptomics and open chromatin data to filter comprehensive prior information about candidate interactions and predict cell-specific gene regulatory network versions using machine learning, while sparse single cell measurements are imputed using local cell similarities. In this way, rare measurements across cell types and a larger condition space for network inference can be exploited, using the natural potential of chromatin accessibility data as a filter to map interactions into a cell-specific context.A distinguishing feature of the proposed method is the characterization of local gene regulatory states, which allows the observation of continuous changes throughout a cell-cell similarity embedding. This will be useful to examine changes during cell differentiation and along gradients in spatial reconstructions, for example of embryonic development. The developed methods will be made available to the community as a computational toolkit to improve the characterization of gene regulation by combining different types of data.
Original text from CORDIS.
Participants
- GENOME RESEARCH LIMITED LBG · SAFFRON WALDENCoordinatorUnited Kingdom
Links
Data: CORDIS, © European Union