Recount Explorer is a Shiny app for reading the recount3 collection of RNA-seq studies. recount3 holds hundreds of thousands of human and mouse samples, and one pipeline processed all of them in the same way, so counts from different studies are comparable.
Reading one of those studies normally means installing Bioconductor, learning the recount3 API, and writing R. This app removes that step. Pick a study, look at it, and take the data with you.
glioblastoma finds
studies whose titles never mention it.recount3 stores one file for each study, and there is no way to ask for less
of it. Downloads run at about 100 KB for each sample, and memory at about
30 bytes for each gene in each sample. The app shows both before you load
anything, and refuses studies over 500 samples so a shared server stays up.
RECOUNT_EXPLORER_MAX_SAMPLES moves that limit.
Most studies are small. Of the 18,998 in the catalog, 18,323 are under 10 MB.
Cite the recount3 paper for the data:
Wilks C, Zheng SC, Chen FY, et al. recount3: summaries and queries for large-scale RNA-seq expression and splicing. Genome Biology 22, 323 (2021). doi:10.1186/s13059-021-02533-6
The Export view writes a script that carries this reference in its header, along with the accession and the package versions you used.
Built by Samuel Bharti.
The app is MIT licensed. The source is open, and contributions are welcome. The recount3 project sets the terms for the data itself.
The study catalog is a snapshot that ships with the app, so the list appears at once and needs no network. Loading a study downloads it from the recount3 servers on a background process, which is why the rest of the app stays usable while it runs.
Computation lives in a Shiny-free layer that is tested on its own.