- redone UMAP calculations after one sample removed bc of metadata issue
- now done via rerunning conos on subset of samples
- did this for a subset of oligos+OPCs too