# 03_QUEST_NEW_DATA

Great, now we have a working recreation of figure3 using real data. How about let's
find some new data sets to display.The processing steps should be 

## 1. Objective and goal

The goal is to analyze new dataset which is the human Cardiac Inflammatory Response. Please use the GEO data set GSE145154 at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE145154. The processing steps should download the dataset, running it through the SCimilarity model to get predictions, and saving the output for the web app, similar to what's done at O2_Quest_FIG3.

## 2. TESTS
Confirm the data that is download is correct.  The data in the data set shoud have cells with > 800 genes and <> 4000 genes.  It should also not inlcude cells with unique molecular identifiers (UMI) > 20000, and exclude cells with mitochondrial gene percentage > 10%.
Confirm the same verifications in 02_Quest_FIG3.md

## 3. Web application

Update the web application to support multiple datasets rather than overwriting files in `web/data/`:
1. **UI Changes**: Add a new dropdown menu in the control bar at the top of the web page to select the active dataset (e.g., "Kidney Atlas" vs "Cardiac Inflammatory Response").
2. **Data Structure**: Store the JSON files for each dataset in separate, dataset-specific subfolders within `web/data/` (e.g., `web/data/kidney/` and `web/data/cardiac/`).
3. **Logic**: When the user changes the dropdown selection, the JavaScript should fetch the appropriate JSON files from the selected subfolder and completely refresh the plots to display the new dataset.
copyright: © 2026 Sonia Timberlake & Ryan Bellmore
license: Proprietary - Authorized Workshop Participants Only
distribution_allowed: false