Research & Publications
My research is currently focused on Graphical Network Modeling of Psychiatric disorders, especially ASD, and its related comorbidities, like depression and anxiety. I use hierarchical clustering and module preservation Analyses, along with Graphical Model estimation, via different techniques, to study these networks of interrelated phenotypical phenomena (Interactomes) to shed light on the structure of psychopathologies and their difference at different states, for example as standalone phenomena compared to coexistence with other pathologies. I am also studying the evolution of these intercalated manifestations (behavioral symptoms) over.
Extensive Research Description
In (Anderson et al., 2015), ASD correlational networks were constructed, based on the “Autism Diagnostic Observation Schedule” (ADOS) items. This was the first Network Analysis of ASD, published in the literature. Among others, it revealed that “Reciprocal Social Interaction” nodes, are the most central behaviors affected in ASD, and are closely related to the “Communication” domain nodes. “Stereotyped Behaviors and Restricted Interests” behaviors were more peripheral. Moreover, Item E3 (Anxiety) was the most peripheral node, in almost all networks. Since targeting the most central nodes in these types of networks are proven to yield better therapeutic results, this study implies targeting “Reciprocal Social Interaction” behaviors would result in good therapeutic results.
Montazeri et al., 2018, further focused on the interaction of anxiety and ASD. Composite scores based on the sum of items from subscales of the “Revised Children’s Anxiety and Depression Scale” (RCADS), (“Social Anxiety”, “Generalized Anxiety”, “Panic Disorder” and “Separation Anxiety”) were included as nodes in a partial correlation network, which also included nodes representing different ASD behaviors. These composite anxiety nodes were consistently peripheral in all networks, further suggesting that anxiety, overall, does not play a central role in the psychopathology of ASD.
Montazeri et al., 2020, focuses on depression-ASD interaction. The ASD group had a higher rate of clinical depression and markedly higher “insomnia” and “restlessness” scores in the regularized partial correlation networks. Depression nodes clustered with the anxiety nodes, but remained separate from autism nodes, in a combined network. Moreover, compared to the controls, “insomnia” and “restlessness” nodes were more central in ASD network. Module preservation analysis showed that depression, in a network composed of depression-, anxiety-, and OCD behaviors, was the least preserved module compared to the controls and hence, “atypical”, in ASD.
Results of the genetic project I contributed to, were published in (Li et al., 2016) and indicated new mutations in the histone modifier PRDM6 were associated with isolated non-syndromic Patent Ductus Arteriosus.
Anxiety; Anxiety Disorders; Depression; Software Design; Computational Biology; Autism Spectrum Disorder