Day 5
Morning
- Demonstration: Agentic Science
- Outlook lecture: Spatial transcriptomics
- Practical work: On the last day, you hopefully have the tools to get started with work on your own datasets. We are available for all questions and support. Alternatively, or if you do not have your own project data yet, you are welcome to choose one of the challenges below (or come up with your own!).
- Option 1: Doublet detection in PBMCs
- Option 2: Download and analyze the mystery dataset
- Analyze the dataset according to the workflow we have practiced, paying attention to the following questions:
- What species and tissue is represented in this dataset?
- How many cells are in there?
- Which cell types can you identify?
- Look at the mitochondrial fraction of reads in this dataset. Is this in line with your expectations for this tissue? What could be special about this dataset?
- Option 3: Choose a dataset from the CELLxGENE Discover data portal that matches your domain interests, and come up with a small research question to answer with it.
- To keep things responsive on a typical laptop, pick a dataset with at most ~20,000 cells (you can sort/filter by cell count on the portal, and subsample further after loading if needed).
Afternoon
- Practical work - continued