On May 7, 2025, 12 trainees selected based on their abstract submissions will each give a three-minute presentation on their research. There will be three flash talks for each of HBHL’s four research themes.
Theme One: Neuroinformatics and Computational Modelling
Dylan Mann-Krzisnik
Status: PhD Institution: McGill University Department: QLS / Computer Sciences Supervisor: Yue Li Poster #107
ECLARE: Multi-teacher contrastive learning via ensemble distillation for diagonal integration of single-cell multi-omic data
Abstract
Integrating multimodal single-cell data, such as scRNA-seq and scATAC-seq, is key for decoding gene regulatory networks but remains challenging due to issues like feature harmonization and limited quantity of paired data. To address these challenges, we introduce ECLARE, a novel framework combining multi-teacher ensemble knowledge distillation with contrastive learning for diagonal integration of single-cell multi-omic data. ECLARE trains teacher models on paired datasets to guide a student model for unpaired data, leveraging a refined contrastive objective and transport-based loss for precise cross-modality alignment. Experiments demonstrate ECLARE’s competitive performance in cell pairing accuracy, multimodal integration and biological structure preservation, indicating that multi-teacher knowledge distillation provides an effective mean to improve a diagonal integration model beyond its zero-shot capabilities. Additionally, we validate ECLARE’s applicability through a case study on major depressive disorder (MDD) data, illustrating its capability to reveal gene regulatory insights from unpaired nuclei. While current results highlight the potential of ensemble distillation in multi-omic analyses, future work will focus on optimizing model complexity, dataset scalability, and exploring applications in diverse multi-omic contexts. ECLARE establishes a robust foundation for biologically informed single-cell data integration, facilitating advanced downstream analyses and scaling multi-omic data for training advanced machine learning models.
Matthew Danyluik
Status: PhD Institution: McGill University Department: Integrated Program in Neuroscience Supervisor: Mallar Chakravarty No poster
Distance-independent functional connectivity preferentially emerges in associative regions during adolescence
Abstract
Whole-brain functional dynamics are thought to be governed by cortical geometry, as predicted by neural field theory, which suggests that the interactions between cell populations can be explained by the distance between them. However, we may intuitively expect that associative regions can increasingly couple with regions beyond their neighbours as their integrative cognitive functions develop. Accordingly, we analyzed resting state functional MRI data from two developmental datasets: the Philadelphia Neurodevelopmental Cohort (PNC, n = 1188, aged 8-23) and the psychopathology-enriched Healthy Brain Network (HBN, n = 1058, aged 5-21). For each participant, we used region-wise generalized additive models to predict functional connectivity from geodesic distance, giving individual R2 maps quantifying the distance-dependence of each region’s functional connectivity profile. We also labelled each connection according to the assignment of its two regions in Mesulam’s cortical hierarchy (sensory, unimodal, associative, or limbic), calculating the average distance-connectivity R2 for each connection type. In both datasets, age effects on the distance-connectivity relationship followed a sensory-associative hierarchy: distance-independence preferentially emerged with age in associative-associative connections, while sensory-sensory connections were well explained by distance at all ages. When correlating regional age effects with annotations from the neuromaps toolbox, we saw that distance independence also developed in regions characterized by lower myelin content and greater temporal autocorrelation, two factors thought to favour prolonged plasticity during adolescence. Together, the ability of associative regions to break the constraint of cortical geometry as the brain matures may relate to their innate capacity for plasticity and high-order function.
Sex-specific developmental gene networks linking risk-taking behaviors in rodents and humans
Abstract
Early adversity increases the risk of psychopathology, including addictive-like behaviors, with emerging evidence of sex-specific effects. However, the underlying mechanisms remain unclear. This study investigated the effects of prenatal adversity on reward-seeking and risk-taking behaviors in rats, identifying sex-specific gene networks in the medial prefrontal cortex (mPFC) associated with these behaviors, that may serve as transcriptomic signatures in humans. Sprague Dawley dams were assigned to either a control (ad libitum diet) or food restriction (FR) group starting on gestational day 10. Offspring were tested at postnatal day 90 for lever-pressing behavior under a fixed-ratio 1 (FR1) reinforcement schedule and in the novelty-suppressed feeding (NSF) to assess reward-risk taking in an anxiogenic environment. Bulk RNA sequencing at P0, P21, and P90 (n=10/group) was analyzed using Weighted Gene Co-expression Network Analysis (WGCNA) and TimeNexus to identify FR-responsive gene subnetworks across development. Human orthologs from these subnetworks were used to derive sex-specific expression-based polygenic scores (ePRS), which were tested for associations with risk-taking behavior in the UK Biobank. FR offspring exhibited increased lever pressing for food reward (p=0.035, n=11-12/group), regardless of sex. However, only FR females showed faster initiation of eating in the NSF (p<0.05, Log-rank Mantel–Cox, n=16-19/group). In humans, the female ePRS was associated with risk-taking behavior in women (p=0.003, B=-0.03, n=208,995), while the male ePRS showed no effect in men (p=0.25, n=179,591). This study demonstrates a novel translational approach to identifying sex-specific brain gene networks that may serve as potential biomarkers of susceptibility to risk-taking behaviours in humans.
Theme Two: Mechanistic Models of Neurodegenerative Disorders
Cyril Bolduc
Status: PhD Institution: McGill University Department: Neurology and Neurosurgery Supervisor: Jean-François Poulin No poster
Ablation and chemogenetic inhibition of molecularly defined midbrain dopamine neuronal subtypes reveal their distinct roles in locomotion
Abstract
Despite advances in delineating the molecular diversity and projection profiles of midbrain dopaminergic (DA) neurons, their functional contribution to locomotion, reward, and motivation remains poorly understood. Here, we applied intersectional ablation and chemogenetic approaches to dissect the distinct roles of calbindin (CALB1+) expressing and non-expressing (CALB1-) DA neurons in locomotion. By engineering CreOFF/FlpON and CreON/FlpON autocleavable caspase3 (taCasp3) viral constructs, we selectively ablated CALB1+ and CALB1- DA neurons in Calb1-ires-Cre;Dat-2A-Flp transgenic mice. Post-ablation, we demonstrated that either the loss of CALB1- or CALB1+ DA neurons led to pronounced deficits in voluntary motor initiation and movement acceleration/deceleration. However, we observed a deficit in locomotor learning only following the ablation of CALB1- DA neurons. We then confirmed these findings by injecting either CreOFF/FlpON and CreON/FlpON AAVs expressing the inhibitory DREADD hM4Di. Acute chemogenetic inhibition of CALB1- DA neurons impaired overall locomotor learning, while inhibition of CALB1+ DA neurons did not affect learning intra-day, but affected day-to-day learning. Furthermore, compared to CALB1- DA neurons, the inhibition of CALB1+ DA neurons further impaired the initiation and amplitude of voluntary movements, as well as the velocity and acceleration/deceleration of motor bouts. Together, this work provides causal insights into the functional roles of molecularly defined DA subtypes in locomotion. Specifying the roles for CALB1+ and CALB1- neurons in locomotion provides insights into mesostriatal circuitry underlying movement and its potential dysregulation in disorders such as Parkinson’s disease.
Guofeng Ye
Status: PhD Institution: McGill University Department: Integrated Program in Neuroscience Supervisor: Marie-Hélène Boudrias Poster #212
Effects of non-invasive brain stimulation on beta oscillations during hand movement in stroke individuals
Abstract
After a stroke, more than half of individuals experience motor impairment, significantly affecting their quality of life. Using electroencephalography (EEG), brain oscillatory features such as Movement-related beta desynchronization (MRBD) have been shown to be associated with motor function. In stroke, lower baseline resting beta power and reduced MRBD were reported in the ipsilesional primary motor cortex (M1), where lower values were associated with greater motor impairment. While non-invasive brain stimulation (NIBS) like transcranial direct current stimulation (tDCS) and transcranial alternating current stimulation (tACS) show promise in modulating MRBD and enhancing motor performance in healthy individuals, their effects on beta oscillatory patterns are yet to be investigated in the context of stroke, which was the goal of this study. EEG signals were collected from 10 stroke individuals and 15 healthy aging subjects. Three types of NIBS were tested in 4 stroke individuals: high-density (HD) 2mA tDCS, 1mA HD 70Hz tACS, or sham NIBS. NIBS was applied over the ipsilesional M1 while participants performed handgrips with the contralateral hand. EEG signals were collected before, as well as 15- and 45-minutes post NIBS and compared to those of the control group. Preliminary results show that stroke individuals had lower baseline beta power (64%) compared to the control group. Anodal tDCS increased the depth of the MRBD (49% post-15min, 96% post-45min), while 70Hz tACS decreased it (-46%, -52%). Understanding how NIBS modulates brain patterns is crucial for developing effective rehabilitation strategies. This study will contribute to developing personalized stimulation protocols post-stroke.
Platon Megagiannis
Status: PhD Institution: McGill University Department: Integrated Program in Neuroscience Supervisor: Yang Zhou Poster #226
Autism-associated CHD8 controls reactive gliosis and neuroinflammation via remodeling chromatin in astrocytes
Abstract
Neuroinflammation plays a central role in the progression of various neurodegenerative disorders, such as in Alzheimer’s disease. Recent evidence points to excessive neuroinflammation contributing to disease exacerbation and pathophysiological deterioration in these disorders. Glial cells, particularly astrocytes and microglia, are major mediators of neuroinflammation, responding to inflammatory stimuli in a process termed reactive gliosis. Contemporary research has highlighted reactive gliosis as a potential therapeutic target for neurodegenerative diseases. Understanding the molecular and cellular mechanisms underlying neuroinflammation is critical for developing strategies to modulate gliosis and improve disease outcomes. In this study, we demonstrate that the ASD-associated chromatin remodeler CHD8 regulates the brain’s inflammatory response through its activity in astrocytes. We show that global deletion of the Chd8 gene in adult mice impairs reactive gliosis. Conditional Chd8 deletion specifically in astrocytes—rather than in microglia—suppresses reactive gliosis by inhibiting astrocyte proliferation and morphological changes in a stab-wound injury model. Furthermore, astrocyte-specific Chd8 ablation reduces LPS-induced neuroinflammation, conferring protective effects in mice. Gene expression profiling and ATAC sequencing reveal that Chd8 loss in astrocytes attenuates neuroinflammation by disrupting astrocyte-microglia interactions and altering metabolic and lipid-related pathways, with these changes reflecting an epigenetic reprogramming of astrocyte function. These findings uncover a novel role of CHD8 in regulating neuroinflammation through astrocytes in the adult brain. Our work suggests a therapeutic potential via targeting CHD8 to mitigate harmful neuroinflammation in neurodegenerative conditions. Currently, we are probing the functions of CHD8 in diverse astrocytic populations and their relevance in neurodegenerative conditions.
Theme Three: Applied Cognitive Neuroscience of Brain
Mohammed Noor
Status: Master’s Institution: McGill University Department: Department of Neurology and Neurosurgery Supervisor: Alyson Fournier Poster #318
A combinatorial strategy targeting multiple microRNA clusters identified in an intraocular inflammation model promotes significant axon regeneration and outgrowth
Abstract
Axons within the central nervous system (CNS) fail to regenerate following injury, largely due to the developmental loss of intrinsic regenerative capacity. Numerous individual gene knockdown events have yielded modest regenerative outcomes. Although combinatorial strategies tend to be more effective, modulating multiple genes simultaneously within the same cell is difficult. As such, we aimed to identify pro-regenerative microRNA (miRNA) clusters that target overlapping genes, thereby enhancing their inhibitory effects and concurrently regulating large gene programs to facilitate neuronal intrinsic outgrowth and regeneration in the Pam3cys pro-regenerative intraocular inflammation model. We employed a multimodal sequencing approach, incorporating both RNA and miRNA sequencing of fluorescently activated cell-sorted retinal ganglion cells (RGCs) following Pam3cys intraocular inflammation. The integration of our sequencing data revealed that the target genes of miR-96 and let-7 clusters were significantly enriched under regenerative conditions compared to the control. Subsequent functional validation utilizing a multi-miRNA construct, comprising both the miR-96 and Let-7 genomic clusters, resulted in substantial axon regeneration and neurite outgrowth both in vitro and in vivo. These results demonstrate that the targeting of miRNA genomic clusters promotes robust intrinsic regeneration in neurons. Our findings reveal a novel synergistic role for miRNA clusters in facilitating regeneration, representing a previously unrecognized axis of miRNA-mediated neuronal repair and offering promising potential therapeutic targets for CNS injuries.
Prakriti Gupta
Status: PhD Institution: The Neuro (Montreal Neurological Institute-Hospital) Department: Integrated Program In Neuroscience Supervisor: Julien Doyon Poster #319
Closed-loop targeted memory stimulation during sleep spindles optimizes motor memory consolidation
Abstract
Sleep plays a critical role in motor memory consolidation by enabling memory trace reactivation during non-rapid eye movement (NREM) sleep spindles. This study provides direct evidence by employing real-time closed-loop auditory stimulation during spindles using the deep learning-based Portiloop EEG device, combined with a Targeted Memory Reactivation (TMR) paradigm. Thirty right-handed young adults completed a modified Motor Sequence Learning (MSL) task involving two distinct 8-element finger sequences, each paired with a specific hand and sound. The sound associated with one sequence was replayed during spindles (TMR condition), while the other sequence served as a control (non-TMR). Participants underwent training, pre- and post-sleep tests in a 3.0T MRI scanner, with overnight auditory stimulation administered in a sleep lab. Behavioral results showed enhanced motor performance post-sleep, with a significantly greater Percentage Change in Duration for Correct Sequences (PCD-CS) in the TMR condition compared to non-TMR. Furthermore, the percentage of stimulated sleep spindles (true positives) positively correlated with the TMR effect (difference in PCD-CS between conditions), indicating that more effective spindle stimulation led to greater performance gains. Notably, stimulation of spindle trains (two or more spindles occurring within 6 seconds) was positively correlated with the TMR effect than isolated (more than 6 seconds apart) spindle stimulation. These findings highlight the critical role of spindle activity in sleep-based memory enhancement and suggest that targeted spindle train stimulation could serve as a non-invasive approach to boost motor memory consolidation, with potential therapeutic applications for age-related or neurodegenerative conditions.
Sarah McCrackin
Status: Postdoctoral Researcher Institution: McGill University Department: Psychology Supervisor: Jelena Ristic No poster
From surgical masks to niqabs: How different face coverings influence social judgments
Abstract
Faces signal important messages about our emotional states. Research shows that emotional perception is disrupted when faces are visually occluded, for instance with surgical face masks. Here, we examined if social context within which face occlusion occurs (i.e., medical, cultural, fashion) affected how emotional messages from faces are evaluated. To do so, we presented 62 North American women with images of female Caucasian and Middle Eastern faces bearing neutral, happy, angry, disgusted, or surprised emotional expressions. The faces were either unoccluded or occluded by a surgical mask and cap (medical occlusion), a niqab (cultural occlusion), or a winter scarf and hat (fashion occlusion), with each condition revealing the exact same upper face region. Participants reported the primary emotion displayed by the face and rated their empathy for each face. As before, both emotion recognition and empathy were impaired by face occlusion, and this was similar regardless of the type of face occlusion. There was also less accurate emotion recognition for Middle Eastern faces than Caucasian ones, alongside lower empathy for disgust and anger. Importantly, the difference in response to Caucasian and Middle Eastern faces was larger when faces were occluded. Together, these results show that while face occlusion exerts overall negative consequences on human social communication, these effects become more pronounced in culturally unfamiliar contexts (i.e., with outgroup bias).
Theme Four: Population Neuroscience and Brain Health
Camille Héguy
Status: PhD Institution: McGill University Department: Integrated Program in Neuroscience Supervisor: Marie Brossard-Racine Poster #403
Very preterm birth and neonatal hypoxic-ischemic encephalopathy distinctly impacts early hippocampal development
Abstract
The most active period of hippocampal growth occurs during the last two trimesters of gestation and continues up until the age of 2 years. Critical early life events [ELE], such as very preterm [VPT] birth (<32 weeks gestational age [GA]) or neonatal hypoxic-ischemic encephalopathy [HIE], coincide with this acute period of hippocampal growth, rendering this brain region vulnerable to developmental disruptions. Therefore, this study aimed to compare the distinct effects of two different ELE on regional hippocampal development. We recruited VPT-born, term-born with moderate to severe HIE treated with therapeutic hypothermia, and healthy control [CTL] neonates. Enrollees completed a brain MRI around term age. Total hippocampal volume and volumes of 7 subfields were extracted using Hippunfold, a validated automatic segmentation pipeline. Regional volume to total brain volume [TBV] ratios were calculated for each hippocampal region bilaterally and group differences were evaluated using ANOVA, correcting for multiple comparisons. Compared to CTL (n=12), VPT neonates (n=38) presented with smaller right CA1 (p<0.01) and CA4/DG (p<0.05) ratios, while HIE neonates (n=28) presented with smaller right CA4/DG ratios (p<0.05), and smaller TBV (p<0.05). Compared to HIE, VPT neonates presented with smaller right CA1 (p<0.05), greater left subiculum (p<0.05) and greater total right hippocampus (p<0.001) ratios. Although both VPT and HIE neonates showed increased vulnerability in the right CA4/DG region, premature extrauterine exposition had a more pronounced effect on right CA1 development at TEA compared to a term hypoxic-ischemic event. In contrast, HIE appears to negatively impact overall brain development.
Farida Zaher
Status: Master’s Institution: McGill University Department: Psychiatry Supervisor: Lena Palaniyappan No poster
Natural Language Processing to detect mental state effects of ketamine in depression: A pilot study
Abstract
Background: In current psychological practices, mental states are mostly detected through self-report. Applications of Natural Language Processing (NLP) to speech in the clinic can help assess the mental states of individuals. This is an objective way to identify subtle changes in mental states and quantify them. Ketamine is also said to be a model for symptoms of psychosis. Positive symptoms of psychosis – such as perceptual disturbances, negative symptoms of psychosis – such as poverty of speech, as well as formal thought disorder can all be produced by the administration of Ketamine. In this ongoing study, we aim to use NLP to detect Ketamine-induced speech changes to provide a means to test: (1) if Ketamine can reduce depression-related features of language, to further studies establishing NLP speech markers as an objective marker of mental states; and (2) the validity of Ketamine as a model of linguistic aspects of psychosis. Methodology: Speech samples will be collected from 20 patients with depression undergoing ketamine treatment at the Douglas Ketamine Clinic. Results: Our results demonstrated that the situation model (SM) numerically improves over time with ketamine treatment. A paired samples t-test revealed a significant difference in their depression scores before and after treatment, which was also associated with the improved situation model over time. Conclusions: This study will help advance the development of objective markers of mental states, furthering the applications of NLP which can contribute to relapse detection, reliable diagnoses of mental illness and testing the efficacy of treatment through language analysis.
Nada Hannoui
Status: Master’s Institution: Douglas Research Institute Program: CRISP Lab Supervisor: Geneviève Sauvé No poster
The impact of the MINDS@WORK program on cognitive complaints: A psychosocial workplace evaluation
Abstract
Individuals living with severe mental illness (SMI), such as schizophrenia, face significant barriers to employment despite their desire to work. While existing employment support program have shown some success in job attainment, they remain insufficient in sustaining long-term employment. Cognitive impairments—such as deficits in attention, memory, and executive functioning—are among the key factors limiting job retention in this population. To address these challenges, our research team developed Minds@Work, an innovative group-based psychosocial intervention integrating cognitive remediation, self-compassion training, and socio-emotional skill development. This study aims to assess its feasibility, acceptability, and preliminary effectiveness in improving job retention and reducing cognitive complaints among individuals with SMI The intervention was delivered by trained professionals, and outcomes were measured using a pragmatic pretest-posttest design. Key variables included cognitive complaints, motivation, workplace integration, and self-reported employment duration. Findings demonstrated promising results, with participants reporting enhanced motivation, improved cognitive functioning, and greater social-emotional skills, alongside longer employment durations. This study contributes to the understanding of cognitive interventions in employment outcomes for individuals with SMI. The positive results suggest that Minds@Work could serve as a scalable, evidence-based solution to improve job retention and workplace functioning for this underserved population. Results shows improvement of cognitive difficulties within different subject as the level of education.
HBHL Symposium 