Add more data to recount3

[lcolladotor]

This is a recurrent goal as new data is deposited nearly every day to the Sequence Read Archive.

• To add more data to recount3, we first need computing credits at some large computing clusters such as ACCESS (formerly called XSEDE) https://access-ci.org/.

• Next, we have to run Monorail https://github.com/langmead-lab/monorail-external to process new data.

• The outputs are then transferred to a local cluster where we can keep a backup of the data. On the recount3 paper, this is called the aggregation node. There files across studies are aggregated.

• The data is then uploaded to IDIES, AWS Open Data Sponsorship Program https://aws.amazon.com/marketplace/pp/prodview-t3rflz3f557jq#resources, AnVIL, or any other active mirrors. It has to follow the data structure that the recount3 R package expects.

There are additional steps that are part of the recount3 world such as:

• generating tissue predictions and all predictions that Shijie C. Zheng ran for the initial recount3 release
• there are processing steps needed to generate the Snaptron compilations. Christopher Wilks ran this for the initial release
• Afrooz Razi et al recently also used part of the Monorail output that is not publicly shared to obtain genotype predictions https://doi.org/10.1101/2023.10.21.562237
• Update the recount3 study browser https://github.com/LieberInstitute/recount3-docs/tree/master/study-explorer

This goal really falls outside the recount3 R package, though the R package is one of the most commonly used interfaces for the data. Accomplishing this goal will likely need its own support and/or coordination with Wilks et al and/or Razi et al