Shiny - Coexpression enhances cross-species integration of single-cell RNA sequencing across diverse
Last updated
Last updated
Single-cell RNA sequencing is increasingly used to investigate cross-species differences driven by gene expression and cell-type composition in plants. However, the frequent expansion of plant gene families due to whole-genome duplications makes identification of one-to-one orthologues difficult, complicating integration. Here we demonstrate that coexpression can be used to trim many-to-many orthology families down to identify one-to-one gene pairs with proxy expression profiles, improving the performance of traditional integration methods and reducing barriers to integration across a diverse array of plant species.
... Integrating cross-species single-cell data is an increasingly common goal in the fields of plant development, evolution and molecular biology. To facilitate this process, we have demonstrated that using coexpression proxies expands the gene space available for integration. To facilitate adoption of this approach by the community, we have generated pairwise coexpression proxies between 13 plant species at 3 thresholds. All coexpression proxy lists have been made available at https://gillislab.shinyapps.io/epiphites_v11/. In addition, we have provided a workflow for generating a coexpression network from scRNA-seq data and using it to identify coexpression proxies for integration (Supplementary Code), which additionally requires only gene phylogenies between the two species. We show that this approach generates networks similar to gold standard networks and enables similar integration (Supplementary Figs. 3 and 4). These proxy lists provide an important resource for improving the integration of single-cell data, accelerating the transfer of knowledge from well-studied model organisms to crop systems that are crucial to the global food supply.
