iiinova — where your digital self keeps evolving

And of course a huge thanks to my postdoc mentors, @stephaniehicks.bsky.social and @martinowk.bsky.social for their support (and patience 😅) with this project!

240

This was a huge team effort, so special shout out to the wet lab biologist who generated all of this data. Namely co-first author @svitlanabach.bsky.social, as well as Madeline Valentine and Sarah Maguire. Also huge thank you to other in the neuropath and neural plasticity teams here at LIBD.

This is a massive dataset and the first of it's kind, so I really was only able to scratch the surface of what insights it can provide. I'm excited to see how the community uses this resource, and for future projects in this space! 12/12

GitHub - LieberInstitute/spatialAmygdala

We additionally developed web-based Vitessce apps for exploring each of these datasets freely on your browser! I'm a big fan, so please check them out! You can find them along with the analysis and links to raw data files on the landing page for our github repo. 11/12 github.com/LieberInstit...

github.com

Contribute to LieberInstitute/spatialAmygdala development by creating an account on GitHub.

141

Finally, we wanted to ask how genetic risk for psychiatric disease localized within the human amygdala. gsMap revealed broad enrichment for brain-related traits in gray matter areas (in line with a highly polygenic architecture) whereas scDRS revealed nuclei specific enrichment. 10/12

And we were able to discover considerable cell type heterogeneity in the central nucleus of the amygdala, revealing spatially distinct populations of PRKCD/PENK/CARPT vs SST/TAC3/CRH neurons that appear to resemble those found in mice CeA. 9/12

140

We next chose to use VisiumHD to provide full-transcriptome profiling at single cell resolution to characterize the intercalated cells and central nucleus. Combing this with reference atlases revealed at least 3 reproducible ITC cell types - two island and one non-island population. 8/12

This revealed that different amygdala nuclei contained completely distinct neuronal populations, and showed within nuclei heterogeneity such as the magnocellular and parvocellular divisions of the basal nucleus. 7/12

Several snRNA-seq references have recently been published describing the cell type diversity in the human amygdala. We were able to take these references (both Yu et al. and our previous dataset) and spatially map them into both our Visium and Xenium datasets. 6/12

Using the Xenium imaging-based technology, we were able to again reconstruct the known amygdala anatomy using a smaller targeted gene panel, while also providing validation for many novel marker genes at single cell resolution. 5/12

Analyzing spatial co-expression networks allowed us to recover gene programs specific to distinct amygdala nuclei and identify core transcription factors that may play a central role in their molecular identity and function. 4/12

140

Doing so allowed us to recover the known anatomy of the amygdala using unsupervised methods. This includes basolateral complex nuclei, medial and central nuclei, the cortical nucleus, and intercalated islands! As well as surrounding structures such as the hippocampus and claustrum. 3/12

Our understanding of the molecular and cellular make up of the human amygdala has lagged far behind other structures. So we used Visium, Xenium, and VisiumHD to comprehensively profile the amygdala at full-transcriptome and single cell resolution across 9 total donors. 2/12

Michael Totty @mictott.bsky.social · 21/08/2026

Very cool work, Jingyi! Did you also perform a mouse-level label permutation for DV, as you did for DE? Is it possible the large differences in number of cells and mice across groups cause between-mouse variation to appear as DV?

011

Michael Totty @mictott.bsky.social · 21/08/2026

Interesting work! Vertes et al tested this in 2007 and found that mPFC->RE->HPC connections do exist via electron microscopy. Any thoughts on how these results might be reconciled? PL vs IL?

Michael Totty @mictott.bsky.social · 03/08/2026

It took 14 months, but I’m incredibly excited to share that my NIMH Pathway to Independence Award (K99/R00) has officially been funded!

8942

Michael Totty @mictott.bsky.social · 03/08/2026

But then the signal flipped under external disinhibition. A novel, loud tone drove a decrease in freezing but also a huge RE transient. I assumed it had to be tracking perceived/emotional salience, so it’s cool to this work out!

110

Michael Totty @mictott.bsky.social · 03/08/2026

Congrats Ben and co! I saw a very similar pattern in the nucleus reuniens, though we never really published or followed up on it. RE photometry signal (not DA) was high at US, high in early extinction, low in late ext. And low at ext retrieval. Seemingly tracking associative strength.