Biology subjects

Brkic, D.

Publications and source records attributed to Brkic, D..

3 recordsLinked to original sources

bioRxiv2023

The impact of ROI extraction method for MEG connectivity estimation: practical recommendations for the study of resting state data.

Magnetoencephalography and electroencephalography (M/EEG) seed-based connectivity analysis requires the extraction of measures from regions of interest (ROI). M/EEG ROI-derived source activity can be treated in different ways. It is possible, for instance, to average each ROIs time series prior to calculating connectivity measures. Alternatively, one can compute connectivity maps for each element of the ROI prior to dimensionality reduction to obtain a single map. The impact of these different strategies on connectivity results is still unclear. Here, we address this question within a large MEG resting state cohort (N=113) and within simulated data. We consider 68 ROIs (Desikan-Kiliany atlas), two measures of connectivity (phase locking value-PLV, and its imaginary counterpart- ciPLV), three frequency bands (theta 4-8 Hz, alpha 9-12 Hz, beta 15-30 Hz). We compare four extraction methods: (i) mean, or (ii) PCA of the activity within the seed or ROI before computing connectivity, map of the (iii) average, or (iv) maximum connectivity after computing connectivity for each element of the seed. Hierarchical clustering in then applied to compare connectivity outputs across multiple strategies, followed by direct contrasts across extraction methods. Finally, the results are validated by using a set of realistic simulations. We show that ROI-based connectivity maps vary remarkably across strategies in terms of connectivity magnitude and spatial distribution. Dimensionality reduction procedures conducted after computing connectivity are more similar to each-other, while PCA before approach is the most dissimilar to other approaches. Although differences across methods are consistent across frequency bands, they are influenced by the connectivity metric and ROI size. Greater differences were observed for ciPLV than PLV, and in larger ROIs. Realistic simulations confirmed that after aggregation procedures are generally more accurate but have lower specificity (higher rate of false positive connections). Though computationally demanding, after dimensionality reduction strategies should be preferred when higher sensitivity is desired. Given the remarkable differences across aggregation procedures, caution is warranted in comparing results across studies applying different methods.

neuroscience↗

bioRxiv2020

Gene functional networks and autism spectrum characteristics in young people with intellectual disability

BackgroundGenes associated with Intellectual disability (ID) can be grouped into networks according to gene function. This study asked whether individuals with ID show differences in autism spectrum characteristics (ASC), depending on the functional network membership of their rare, pathogenic de novo genetic variants. MethodsChildren and young people with ID of known genetic origin were allocated to two broad functional network groups: synaptic physiology (n=29) or chromatin regulation (n=23). We applied principle components analysis to the Social Responsiveness Scale to map the structure of ASC in this population, and identified three components - Inflexibility, Social Understanding and Social Motivation. We then used Akaike Information Criterion (AIC) to test the best fitting models for predicting ASC components, including demographic factors (age, gender), non-ASC behavioural factors (global adaptive function, anxiety, hyperactivity, inattention) and gene functional networks. ResultsWe found that, when other factors are accounted for, the chromatin regulation group showed higher levels of Inflexibility. We also observed contrasting predictors of ASC within each network group. Within the chromatin regulation group, Social Understanding was associated with inattention, and Social Motivation was predicted by hyperactivity. Within the synaptic group, Social Understanding was associated with hyperactivity, and Social Motivation was linked to anxiety. ConclusionWe report that gene functional networks can predict Inflexibility, but not other ASC dimensions. Contrasting behavioural associations within each group suggests network-specific developmental pathways from genomic variation to autism. Simple classification of neurodevelopmental disorder genes as high risk or low risk for autism is unlikely to be valid or useful.

genetics↗

bioRxiv2018

Functional network dynamics in a neurodevelopmental disorder of known genetic origin

Cognitive processing depends on the temporal co-ordination of functional brain networks. This fundamental aspect of neurophysiology potentially bridges the genetic regulation of neuronal activity and developmental cognitive impairments. We investigated brain network dynamics in a neurodevelopmental disorder of known genetic origin, by comparing individuals with ZDHHC9-associated intellectual disability to individuals with no known impairment. We used Hidden Markov Modelling on magnetoencephalography (MEG) data, at rest and during auditory oddball stimulation, to characterise transient network dynamics. At rest, network dynamics distinguished the groups, with ZDHHC9 participants showing longer state activation. Crucially, ZDHHC9 gene expression levels predicted the group differences across networks, supporting an association between molecular pathology and neurophysiology. In contrast, network dynamics during auditory oddball stimulation did not show this association. We demonstrate a link between brain network dynamics and regional gene expression, and present a valuable method for understanding the real-time neural mechanisms linking genetic variation to cognitive difficulties.

neuroscience↗