Michael Totty
@mictott.bsky.social
Postdoctoral neuroscientist at JHU working at the intersection of neural circuits, bioinformatics, and psychiatric disorders.
https://mictott.github.io
https://mictott.github.io
Finally, leveraging the human dataset, we conducted cell type enrichment for heritability of psychiatric traits (GWAS), as well as MDD and PTSD-associated transcriptomic changes. Aligning with previous literature, we found that GABAergic neurons were enriched for both. 11/13
September 18, 2025 at 1:10 AM
Finally, leveraging the human dataset, we conducted cell type enrichment for heritability of psychiatric traits (GWAS), as well as MDD and PTSD-associated transcriptomic changes. Aligning with previous literature, we found that GABAergic neurons were enriched for both. 11/13
Of importance, using the Yu et al dataset, we found amydala neuron types were markedly less conserved from humans to mice than humans to NHPs. While one to one cell type mapping could still be observed for inhibitory neurons, this was not true for excitatory cells. 10/13
September 18, 2025 at 1:10 AM
Of importance, using the Yu et al dataset, we found amydala neuron types were markedly less conserved from humans to mice than humans to NHPs. While one to one cell type mapping could still be observed for inhibitory neurons, this was not true for excitatory cells. 10/13
Leveraging the precise subregion punches in NHPs, pseudobulk differential analysis revealed highly specific marker genes for excitatory neurons found in the LA, BA, and aBA across species. We confirmed these novel markers using in situ hybridization in both humans and macaques. 9/13
September 18, 2025 at 1:10 AM
Leveraging the precise subregion punches in NHPs, pseudobulk differential analysis revealed highly specific marker genes for excitatory neurons found in the LA, BA, and aBA across species. We confirmed these novel markers using in situ hybridization in both humans and macaques. 9/13
Moving onto excitatory neurons, we found that the LA, BA, and aBA subnuclei are almost exclusively composed of distinct classes of excitatory neuron types, and these neurons display relatively strong cross-species conservation (though less than inhibitory). 8/13
September 18, 2025 at 1:10 AM
Moving onto excitatory neurons, we found that the LA, BA, and aBA subnuclei are almost exclusively composed of distinct classes of excitatory neuron types, and these neurons display relatively strong cross-species conservation (though less than inhibitory). 8/13
Within the BLA, we additionally found the aBA contains a large number of CCK+/CNR1+ interneurons, and that the LA, BA, and aBA nuclei were enriched for distinct subpopulations of SST+ and PVALB+ neurons. 7/13
September 18, 2025 at 1:10 AM
Within the BLA, we additionally found the aBA contains a large number of CCK+/CNR1+ interneurons, and that the LA, BA, and aBA nuclei were enriched for distinct subpopulations of SST+ and PVALB+ neurons. 7/13
Compositional analyses using linear mixed models (crumblr) found that the central nucleus was enriched in neurons derived from the LGE (PRKCD+, SST/TAC1+, DLK1+, etc), whereas the BLA is enriched in MGE (SST+,PVALB+) and CGE (LAMP5+, CCK+) cell types. 6/13
September 18, 2025 at 1:10 AM
Compositional analyses using linear mixed models (crumblr) found that the central nucleus was enriched in neurons derived from the LGE (PRKCD+, SST/TAC1+, DLK1+, etc), whereas the BLA is enriched in MGE (SST+,PVALB+) and CGE (LAMP5+, CCK+) cell types. 6/13
For inhibitory neurons, we found a wide variety of cell types (putatively) spanning medial, caudal, and lateral ganglionic origins - including two intercalated cell types (TSHZ1+) - which all displayed strong cross-species conservation. 5/13
September 18, 2025 at 1:10 AM
For inhibitory neurons, we found a wide variety of cell types (putatively) spanning medial, caudal, and lateral ganglionic origins - including two intercalated cell types (TSHZ1+) - which all displayed strong cross-species conservation. 5/13
This resulted in ~130,000 single nuclei across the three species, capturing a wide diversity of inhibitory and excitatory neurons, as well as non-neuronal populations. 4/13
September 18, 2025 at 1:10 AM
This resulted in ~130,000 single nuclei across the three species, capturing a wide diversity of inhibitory and excitatory neurons, as well as non-neuronal populations. 4/13
Despite the difficulties working with fresh frozen postmortem brains, @svitlanabach.bsky.social
was able to precisely capture the basolateral complex from 5 human donors. We then performed cross-species integration to decode subnuclei origins of human amygdala cell types. 3/13
was able to precisely capture the basolateral complex from 5 human donors. We then performed cross-species integration to decode subnuclei origins of human amygdala cell types. 3/13
September 18, 2025 at 1:10 AM
Despite the difficulties working with fresh frozen postmortem brains, @svitlanabach.bsky.social
was able to precisely capture the basolateral complex from 5 human donors. We then performed cross-species integration to decode subnuclei origins of human amygdala cell types. 3/13
was able to precisely capture the basolateral complex from 5 human donors. We then performed cross-species integration to decode subnuclei origins of human amygdala cell types. 3/13
In a multi-institute effort, we conducted snRNA-seq of the baboon, macaque, and human amygdala. In baboons and macaques, @vincentcostaphd.bsky.social and his lab were able to take precise punches of the lateral (LA), basal (BA), accessory basal (aBA), and central nuclei (CeA). 2/13
September 18, 2025 at 1:10 AM
In a multi-institute effort, we conducted snRNA-seq of the baboon, macaque, and human amygdala. In baboons and macaques, @vincentcostaphd.bsky.social and his lab were able to take precise punches of the lateral (LA), basal (BA), accessory basal (aBA), and central nuclei (CeA). 2/13
Wow, what a thoughtful care package from @cziscience.bsky.social. Excited to join the Faculty Application Bootcamp to prep for the upcoming job cycle!
July 30, 2025 at 7:35 PM
Wow, what a thoughtful care package from @cziscience.bsky.social. Excited to join the Faculty Application Bootcamp to prep for the upcoming job cycle!
SpotSweeper is able to accurately identify these artifacts using the multiscale local variance of mito %. Removal of these hangnails improves downstream results, such as marker gene detection.
June 6, 2025 at 12:03 PM
SpotSweeper is able to accurately identify these artifacts using the multiscale local variance of mito %. Removal of these hangnails improves downstream results, such as marker gene detection.
For example, unwanted tissue hangnails are often created during tissue collection from large samples, such as postmortem human brain. We found that these hangnails present as areas low mito % variance and miscluster with cortical layers.
June 6, 2025 at 12:03 PM
For example, unwanted tissue hangnails are often created during tissue collection from large samples, such as postmortem human brain. We found that these hangnails present as areas low mito % variance and miscluster with cortical layers.
SpotSweeper also addresses a second big problem in spatial transcriptomics: the detection of regional artifacts that may arise due to sample prep.
June 6, 2025 at 12:03 PM
SpotSweeper also addresses a second big problem in spatial transcriptomics: the detection of regional artifacts that may arise due to sample prep.
A big update from the preprint, we now show that SpotSweeper works across tissue types and sequencing-based technologies.
June 6, 2025 at 12:03 PM
A big update from the preprint, we now show that SpotSweeper works across tissue types and sequencing-based technologies.
We also found a set of systematically low quality spots in the Visium platform, and provide evidence that this likely stems from the underlying synthetic barcodes that lead to downstream PCR biases.
June 6, 2025 at 12:03 PM
We also found a set of systematically low quality spots in the Visium platform, and provide evidence that this likely stems from the underlying synthetic barcodes that lead to downstream PCR biases.
Using this approach, we show that SpotSweeper successfully retains high quality observations while accurately detecting problematic local outliers in human DLPFC samples.
June 6, 2025 at 12:03 PM
Using this approach, we show that SpotSweeper successfully retains high quality observations while accurately detecting problematic local outliers in human DLPFC samples.
Like the game minesweeper, SpotSweeper overcomes this limitation by using the information from nearest neighbors to flag problematic local outliers 😎
June 6, 2025 at 12:03 PM
Like the game minesweeper, SpotSweeper overcomes this limitation by using the information from nearest neighbors to flag problematic local outliers 😎
Standard QC metrics (library size, gene count, mito %) confound biology in spatial data. Applying snRNA-seq QC removes spots in regions with inherently sparse RNA — like white matter or cortical L1 — leading to biased results.
June 6, 2025 at 12:03 PM
Standard QC metrics (library size, gene count, mito %) confound biology in spatial data. Applying snRNA-seq QC removes spots in regions with inherently sparse RNA — like white matter or cortical L1 — leading to biased results.
Finally, we found that the RE contains a state signal that strongly encodes the encoding defensive state of the animal. RE activity is high during non-freezing and dramatically quietens during defensive freezing behavior.
April 7, 2025 at 12:46 PM
Finally, we found that the RE contains a state signal that strongly encodes the encoding defensive state of the animal. RE activity is high during non-freezing and dramatically quietens during defensive freezing behavior.
Expanding on this, we used both fiber photometry and single unit ephys to see if the rodent RE also tracks the associative value of the CS. While we found that RE bulk activity tracks associative value, we also found that difference ensembles of neurons showed distinct coding of fear vs extinction.
April 7, 2025 at 12:46 PM
Expanding on this, we used both fiber photometry and single unit ephys to see if the rodent RE also tracks the associative value of the CS. While we found that RE bulk activity tracks associative value, we also found that difference ensembles of neurons showed distinct coding of fear vs extinction.
Co-lead Muhammad Badarnee of Milad's group re-analyzed some of their existing data to find that the human RE appears to both track the associative value of conditioned stimuli and synchronizes with the frontal cortex, hippocampus, and amygdala. This appears to nicely match our findings in rodents.
April 7, 2025 at 12:46 PM
Co-lead Muhammad Badarnee of Milad's group re-analyzed some of their existing data to find that the human RE appears to both track the associative value of conditioned stimuli and synchronizes with the frontal cortex, hippocampus, and amygdala. This appears to nicely match our findings in rodents.
First time in Vanderbilt’s Memorial Gymnasium in 8 years. I grew up with my dad driving us 1.5 hours to probably 8+ games a year. It’s good to be back!
December 31, 2024 at 12:55 AM
First time in Vanderbilt’s Memorial Gymnasium in 8 years. I grew up with my dad driving us 1.5 hours to probably 8+ games a year. It’s good to be back!
What an incredible week attending ACNP for the first time! Honored that I had the opportunity to present my latest work as a Data Blitz talk. Already looking forward to next year’s meeting! #ACNP2024
December 14, 2024 at 1:40 AM
What an incredible week attending ACNP for the first time! Honored that I had the opportunity to present my latest work as a Data Blitz talk. Already looking forward to next year’s meeting! #ACNP2024