Cognitive and genetic mechanisms underlying fear learning and extinction

To apply for this project please visit the LIDo website: https://www.lido-dtp.ac.uk/apply

Aim. Examine cognitive and genetic mechanisms underlying fear learning and extinction and their links with behavioural avoidance.

Rationale. Fear is an evolutionarily useful emotion. One mechanism through which it develops is associative learning, as modelled in fear conditioning paradigms which include two core phases. In fear acquisition, a neutral stimulus (e.g. a shape) is repeatedly paired with an aversive stimulus (e.g. harsh sound). Through associative learning, this leads to fear development. In extinction, the conditioned stimulus is repeatedly shown with no aversive element allowing new safety learning to take place (updating of associative learning). Fear learning is commonly associated with behavioural avoidance, precluding the opportunity for fear extinction.

Twin and genomic studies show genetic influence on the mental processes underlying fears. Computational studies have begun to reveal reliable translational concomitants of these mental processes. However, little is known about the specific psychological processes through which genetic risk on these mental processes is mediated. To date, genetic and experimental fear studies have been independent. We will combine twin and genome-wide methods with an experimental fear conditioning measure. We will use computational modelling to examine causal pathways from genetic risk to fear learning and extinction, uncovering the genetic mechanisms that contribute towards the mental processes (e.g. associative learning) supporting fear. This is very timely given recent huge growth in genome-wide studies of fear-related phenotypes.

Prep work. The first supervisor co-leads a study of ~40k recontactable participants with genome-wide array data. The first supervisor and partner have collaboratively built a smartphone app which remotely delivers a fear conditioning task. App data has been collected from >2,750 individuals including ~500 twin pairs. This data would benefit from computational learning models, which more sensitively decompose fear conditioning into component mental processes. These are complex and require specific expertise. The second supervisor’s team have extensive experience applying computational approaches to this type of data. Joint exploratory analyses have been undertaken, demonstrating the viability of this collaboration.

Aim 1. Using pre-existing twin data, build and test computational models (with 2nd supervisor) to examine mental processes underlying fear conditioning.

Aim 2. Conduct twin analyses to examine the extent to which these fear processes, self-reported fear, and behavioural avoidance are associated with the same/different genetic factors.

Aim 3. Using newly collected data (N~1,000) replicate computational analyses from Aim 1. 

Aim 4. Examine the role of whole genome genetic risk in the associations between fear acquisition, extinction and behavioural avoidance. 

Importance. Fear and avoidance influence educational attainment and wider life goals. These behaviours confer both advantage and disadvantage at times. Identifying the mechanisms underlying individual differences in these traits is key to understanding how and why fears develop and would support innovation in multiple areas.

Academic Supervision. Prof Eley has a large active team and provides fortnightly supervision. Through weekly meetings, all team members learn about one another’s work enabling considerable peer support and collaboration. She is based in the SGDP Centre, IoPPN, KCL, where there is strong expertise in the analysis of twin and genomic data, as well as fear conditioning data. Prof Robinson has a large team within the Neuroscience and Mental Health group at the Institute of Cognitive Neuroscience, UCL, who are experts in Computational Modelling. The student will join the regular group meetings, journal clubs and invited talks as well as the wider departmental and school environment. The UCL and KCL expertise are different but complimentary, and this interdisciplinary project will not be possible without each. The student will draw on both formal and peer training within both departments and will also attend courses hosted both within and outside the UK.

Industrial Partner: Torchbox is an established, employee-owned company providing software engineering and app development expertise for the not-for-profit and public sectors. A large proportion of their portfolio of clients are universities and other research institutions, including the first supervisor’s team at KCL. Many of the staff come from an academic background. Torchbox has a mature programme for managing interns and offers a variety of approaches to training suitable for the student. They would likely begin with shadowing experienced team members. The primary skill set the student would learn relates to overall management of projects, as this is a particular strength of our approach. Depending on their prior expertise and particular interests, the student may also be able to gain training in data modelling, requirements gathering, user research/user needs analysis, and even potentially programming.

Lab Culture. The KCL EDIT Lab, the UCL Neuroscience and Mental Health group and Torchbox teams strive to provide diverse environments that are open, welcoming, and supportive to all. KCL-wide initiatives are supported by an active Diversity & Inclusion Team. These groups are currently predominately made up of people from white backgrounds, and the EDIT lab is predominantly female. To ensure that we are the diverse teams we strive to be, we particularly welcome applications from underrepresented group, including individuals from any other race or ethnicity background (E.g., Black, Asian, Latinx). Please see the EDIT Lab group culture tab for more information on our general approach to this.