Treehouse Data Compendia

A group at UCSC (Treehouse: https://treehousegenomics.soe.ucsc.edu/ ) has compiled a large repository of controlled access RNA-seq data. We've started discussions around how we could short circuit the process of acquiring controlled access data ourselves by teaming up with them.

One path that might be reasonable:

• Provide a simple stub server that can share files + metadata from some standard format in a directory (this would allow treehouse, a core with their own data, or other organizations to provide a source repository)
• Write a surveyor for the simple server.
• Implement functionality that allows a data refinery instance to look at another instance to snag quantiles for platforms. This provides a more federated system where individual sites could run their own data refineries.
• Spin up a data refinery at the UCSC treehouse that snags quantiles and surveys their controlled access data. The final outputs (summarized gene expression values) would be processed in the manner ours are and also should not be controlled access at that stage.

Review the Treehouse site and the four proposed paths: a stub server, a surveyor, cross-instance quantile retrieval, and a UCSC Treehouse data refinery. Before implementation, clarify which path is in scope, the standard format and interfaces, and what successful federated outputs should contain.