LieberInstitute / spatialDLPFC

spatialDLPFC project involving Visium (n = 30), Visium SPG (n = 4) and snRNA-seq (n = 19) samples
http://research.libd.org/spatialDLPFC/
17 stars 3 forks source link

Investigate "bad" clusters of spots from spatial registration #58

Closed abspangler13 closed 2 years ago

abspangler13 commented 2 years ago

Stemming from Maddy's question in spatial meeting on 03/17/22. Is it always the same spots in the bad cluster? Are they neuropil spots? It's important to decide what these clusters are so we know if we need to drop them from the differential expression analysis.

Notes from meeting with Leo about this on 03/16/22: Make orange and grey spot plots like we did for the data QC so we can see if it's the same spots always in the bad cluster. Make heatmap of comparing number of spots in k = 9 clusters to k = 28 clusters. This will be panel d of figure 2. Use table() to do this. Use pheatmap (or complex heatmap) to make the heatmap. Louise can help if I need.

Other resources for making heatmaps sent to me by Leo: http://research.libd.org/SPEAQeasy-example/bootcamp_intro#52_Visualizing_DEGs https://github.com/LieberInstitute/Visium_IF_AD/blob/master/code/08_harmony_BayesSpace/04_heatmaps_dlpfc_markers.R#L131-L149