Donor mapping for MERFISH MEC/HPF specimens

,
1

I’m working with two SEA-AD MERFISH files from the spatial-transcriptomics bucket: medial-entorhinal-cortex/combined_anndata_object/1444201261_MEC_mapped.h5ad andhippocampus/combined_anndata_object/1444211893_HPF_mapped.h5ad.

The obs in these combined objects contains only cell-type annotations and spatial coordinates — no Donor ID / Specimen Barcode / Section columns like the MTG MERFISH release has, and there’s no per-donor folder in the S3 path (unlike the caudate region). How do I map these specimens/cells to their donor(s) (H20.33.xxx)? Specifically: is each combined file a single donor, or are multiple donors pooled — and if pooled, where are the per-cell donor labels?

Thanks!

2

Hi @Tamarav,

Thanks for reaching out, and for your interest in our data! Each of these .h5ad files represents mapping for a single specimen/section. For each of these files, the Specimen Barcode is in the name - 1444201261 for 1444201261_MEC_mapped.h5ad, for example. However, the Donor ID for both of these files is H24.30.005 - so same donor for both files, but different sections (the MEC_mapped section comes from the MEC, etc).

1 Like
3

Hi @egelfan2 !

Thank you — that resolved the specimen→donor mapping perfectly (barcode 1444201261 = MEC section, 1444211893 = HPF section, both donor H24.30.005).

I’m testing whether a cell-level phenotype I see in the entorhinal cortex tracks Alzheimer’s neuropathology, so it would be helpful to have the donor-level staging for the MERFISH data. Two things I couldn’t resolve on my end:

  1. Neuropathology + demographics for the MERFISH donors. The .h5ad files carry no pathology fields, and H24.30.005 isn’t in the SEA-AD MTG cohort metadata (the 84-donor .33. set) or in the Multiregion 10x donor set (81 donors, also .33.) — so the MERFISH donors appear to be a separate cohort. Is there a donor metadata table for the Multiregion MERFISH release giving, per donor: Braak NFT stage, Thal phase, CERAD score, overall ADNC, APOE genotype, age at death, sex, and cognitive/dementia status?

  2. Cohort composition. How many donors (and sections/regions per donor) are in the Multiregion MERFISH release, and what pathology range does it span? Specifically, does it include both low-/no-pathology (control) and high-pathology (AD) donors — enough to compare across Braak stages?

If there’s a downloadable specimen-barcode → donor-ID → neuropathology crosswalk (analogous to the MTG cohort spreadsheet), that would be ideal — I can join everything from that myself.

Thanks a lot!
Tamara

4

Hi @TamaraV,

I will check with our spatial team on available metadata for this donor. We intended only to profile the MEC and HIP in a neurotypical context, so we could localize all of the transcriptionally defined types we found in the SEA-AD single nucleus -omics datasets from these regions. As a result, this donor is unlikely to have significant pathological burden.

We have not yet generated MEC and HIP data in aged individuals affected different levels of pathological burden.

Best, Kyle

5

Hi @kyle.travaglini ,

Thank you for your reply! I understand. It would be great if you could get more information on available metadata from the spatial team for this donor.

Best,
Tamara