Biology subjects

Dougherty, M.

Publications and source records attributed to Dougherty, M..

6 recordsLinked to original sources

bioRxiv2025

On a roll: Recent familiarity primes the brain to retrieve other memories via dopaminergic nuclei mechanisms

Memory retrieval is notoriously variable. Various neurocognitive states have been theorized to affect retrieval success from moment to moment, but the presence and catalysts of these states in the human brain remain largely understudied. Building on previous work, we studied how recent memory judgments during an ongoing retrieval task (i.e. retrieval judgments vs. novelty detection) and their corresponding neural activity prepare the brain to reinstate unrelated memories during upcoming trials. High-level ventral stream regions, including the hippocampus, reinstated memories with significantly greater fidelity following recent retrieval judgments compared to novelty detection. Activity in dopaminergic nuclei rose during retrieval judgments, which predicted and partially mediated the effect of retrieval judgments on upcoming reinstatement. While dopaminergic nuclei activity predicted upcoming reinstatement, it did not predict upcoming retrieval accuracy. Exploratory analyses revealed the opposite effect in the dorsal and lateral prefrontal cortex, whose activity predicted upcoming retrieval accuracy, but not reinstatement. These results point to distinct neural contributions to what is reinstated and how it may guide memory decisions, and by identifying dopaminergic nuclei as partial mediators of reinstatement, our results open new avenues for investigating how neuromodulatory states may dynamically shape memory accessibility. Significance StatementWhy can we effortlessly recall memories in some moments but struggle at others? Here, we draw on computational models to uncover why human brains are sometimes better prepared to remember and how to nudge them into that state. We discover that recent retrieval judgments, compared to recent novelty detection, increase upcoming accuracy and neural reinstatement of unrelated memories. This effect is so powerful that key memory regions only show reinstatement following preceding retrieval judgments. We found that dopaminergic nuclei are more active during retrieval judgments and predict upcoming memory reinstatement. This pattern partially explains why engaging in remembering prepares your brain to reinstate other memories and reveals new insights for the role dopaminergic nuclei may play in retrieval.

neuroscience↗

bioRxiv2022

Theoretical design of a space bioprocessing system to produce recombinant proteins

Space-based biomanufacturing has the potential to improve the sustainability of deep space exploration. To advance biomanufacturing, bioprocessing systems need to be developed for space applications. Here, commercial technologies were assessed to design space bioprocessing systems to supply a liquid amine carbon dioxide scrubber with active carbonic anhydrase produced recombinantly. Design workflows encompassed biomass dewatering of 1 L Escherichia coli cultures through to recombinant protein purification. Equivalent system mass (ESM) analyses had limited utility for selecting specific technologies. Instead, bioprocessing system designs focused on minimizing complexity and enabling system versatility. Three designs that differed in biomass dewatering and protein purification approaches had nearly equivalent ESM of 357-522 kg eq. Values from the system complexity metric (SCM), technology readiness level (TRL), and degree of crew assistance metric identified a simpler, less costly, and easier to operate design for automated biomass dewatering, cell lysis, and protein affinity purification.

bioengineering↗

bioRxiv2022

Outcome of Crash Course Training on Protein Structure Prediction with Artificial Intelligence

Protein structure predictions have broad impact on several science disciplines such as biology, bioengineering, and medical science. AlphaFold2[1] and RoseTTAFold[2] are the current state-of-the-art AI methods to predict the structures of proteins with an accuracy comparable to lower-resolution experimental methods. In its 2021 year review, both these methods were recognized as "breakthrough of the year" by Science magazine[3] and "method of the year" by Nature magazine [4]. It is timely and important to provide training and support on these emerging methods. Our crash course "Enabling Protein Structure Prediction with Artificial Intelligence "was conducted in collaboration with domain experts and research computing professionals. The crash course was well received by the community as there were 750 registrants from all over the world. Here we provide the summary of the crash course, describe our findings in organizing the crash course, and explain what preparation steps helped us with the hands-on training. CCS CONCEPTSComputing methodologies a Machine learning a Machine learning approaches a Bio-inspired approaches

scientific communication and education↗

bioRxiv2022

Cortical surface electrical potentials are composed of multiple bandlimited frequency components, including high-gamma

A common challenge in neuroscience is how to decompose noisy, multi-source signals measured in experiments into biophysically interpretable components. Analysis of cortical surface electrical potentials (CSEPs) measured using electrocorticography arrays (ECoG) typifies this problem. We hypothesized that high frequency (70-1,000 Hz) CSEPs are composed of broadband (i.e., power-law) and bandlimited components with potentially differing biophysical origins. In particular, the high-gamma band (70-150 Hz) has been shown to be highly predictive for encoding and decoding behaviors and stimuli. Despite its demonstrated importance, whether high-gamma is composed of a bandlimited signal is poorly understood. To address this gap, we recorded CSEPs from rat auditory cortex and demonstrate that the evoked CSEPs are composed of multiple distinct frequency components, including high-gamma. We then show, using a novel robust regression method, that at fast timescales and on single trials during speech production, human high-gamma amplitude cannot be explained by a modulating power-law component; thus, high-gamma is band-limited. Furthermore, we show that the power-law component is less predictive of produced speech compared to the raw high-gamma amplitude. Finally, we show that the largest variance component of human ECoG signals is low-frequency and band-limited, not broadband. Together these results demonstrate that there are multiple, band-limited components of high frequency power in cortical surface electrical potentials, including the high-gamma band, which may have different biophysical origins. Significance StatementElectrocortigraphy (ECoG) records cortical surface electrical potentials (CSEPs). ECoG is utilized in both humans and animal models to understand distributed cortical processing and for brain machine interfaces. The spectral structure of evoked CSEPs is greatly debated. To address this issue, we recorded from rat auditory cortex using {micro}ECoG and human ventral sensory-motor cortex during speech production with high-density ECoG. Using novel analytic approaches, we found that evoked CSEPs are composed of multiple band-limited components, including high-gamma. These results contrast with the dominant thinking in the field that the high-frequency power in ECoG is broadband. Our results raise the possibility that distinct frequency components of ECoG are biomarkers of processing in different cortical layers.

neuroscience↗

bioRxiv2021

High density single-unit human cortical recordings using the Neuropixels probe

A fundamental unit of neural computation is the action potential. While significant advances have been made in the ability to sample action potentials of large numbers of individual neurons in animal models, translation of these methodologies to humans has been lacking due to clinical time constraints, electrical noise in the operating room, and reliability of the methodology. Here we present a reliable method for intraoperative recording of dozens of neurons in humans using the Neuropixels probe, yielding up to [~]100 simultaneously-recorded single-units (n=596 across 11 recordings in 8 participants). Most single-units were active within 1 minute of reaching target depth, compatible with clinical time constraints. Cell pairs active close in time were spatially closer in most recordings, demonstrating the power to resolve complex cortical dynamics. Altogether, this approach provides access to population single-unit activity across the depth of human neocortex at scales previously only accessible in animal models. HighlightsO_LISingle units in 8 patients, yielding 596 putative single units across 11 recordings C_LIO_LIThe majority of putative neurons fire at least 1 spike within one minute after reaching target depth C_LIO_LIPutative neurons take longer to fire at least one spike in anesthetized compared to awake participants C_LIO_LICell-pairs that fire action potentials close in time are spatially closer together than those fire further apart in time C_LI

neuroscience↗

bioRxiv2019

NWB:N 2.0: An Accessible Data Standard for Neurophysiology

Neurodata Without Borders: Neurophysiology (NWB:N) is a data standard for neurophysiology, providing neuroscientists with a common standard to share, archive, use, and build common analysis tools for neurophysiology data. With NWB:N version 2.0 (NWB:N 2.0) we made significant advances towards creating a usable standard, software ecosystem, and vibrant community for standardizing neurophysiology data. In this manuscript we focus in particular on the NWB:N data standard schema and present advances towards creating an accessible data standard for neurophysiology.

neuroscience↗