UCSC-Treehouse/CARE ? reverse-engineered prompt

Build me a simple reproducible pipeline for RNA expression outlier analysis like the Treehouse CARE project.

I want it to run from a single command, ideally with Docker and a Makefile, and use a set of Jupyter notebooks to compare one focus sample against a background compendium of RNA seq expression data. It should take the expression output from an existing RNA seq pipeline, pull down any needed reference files and example data automatically if they are missing, and then run the analysis end to end.

At the end, it should generate two easy to read HTML reports that summarize the results and leave space for adding the sample’s clinical information. Please make it straightforward to test with a sample dataset, and include clear setup and run instructions so someone can clone the repo and get it working without having to figure out a lot of manual steps. If you need to check current docs or tooling details, look them up online first.