Find DLPFC layers

[lcolladotor]

• Try out with BayesSpace
• Could use our layer marker genes from our pilot paper to assign layers
• Maybe consider using our published data and this data when running some clustering algorithms, to then assign layers based on the overlap of the clusters with our initial layer assignments in the published data

Are plaques and tangles enriched in some DLPFC layers?

[lcolladotor]

I think this is related to the following part in the Google Doc:

(Optional) Characterization of laminar distribution of plaques and tangles or those in white matter

[lcolladotor]

@kmartinow suggested another method that might be better than BayesSpace that's a pre-print for a python package

[lcolladotor]

Here's the RESEPT pre-print https://www.biorxiv.org/content/10.1101/2021.07.08.451210v2

[lcolladotor]

From Maddy @madhavitippani: maybe we can use Tangram to find predicted gene expression for the marker genes from the 2021 paper. See https://github.com/broadinstitute/Tangram/blob/master/tangram_tutorial.ipynb

Then use the predicted gene expression data as input for clustering for finding the layers. SCH: it'd be like imputation of the data and then use that for all/most downstream analyses.

[lcolladotor]

Keri: maybe we can use the targeted sequencing data to find the layers if the sparsity from the regular Visium is limiting what methods we can use (like Tangram).

So check if our snRNA-seq marker genes (Matt's and/or Louise's) are in the list of the ~1k genes from the targeted sequencing data.

Could also potentially do a new targeted sequencing run using the snRNA-seq cell type marker genes (if it's feasible in terms of $$ :P)