STRATIFY: Brain network-based stratification of reinforcement-related mental health disorders.
To start: 1 October 2019
Award(s): 3-year fully funded PhD studentship
Project:
Our main body of work has been concerned with the identification of neural mechanisms of reinforcement-related disorders. Here we have made seminal contributions to the field: notably, we comprehensively characterised brain activity during reward processing, including linking a novel genetic risk factor for substance abuse [8]. We also investigated the relation of reward processing to different reinforcement domains and discovered that the ventral striatum is activated by social-emotional stimuli. This activation is dependent on an oxytocin receptor genotype, indicating that oxytocin might regulate the affective valence (hedonic value) of social cue in humans [9, 10]. Decreased ventral striatal activation following social cues in carriers of the “risk” genotype resulted in greater resilience against negative life events, while leading to more social-affective problems under favourable environmental conditions. Charting the integrative nature of neural processes and observing its correlation with nuanced behaviour, we progressed from analyses focusing on specific brain regions of interest to multivariate studies measuring coordinated brain activity. We thereby identified factors integrating brain activation during behavioural control which were differentially associated with substance abuse [11]. Our work has significantly influenced this area of research: sine late 2010, we have published over 50 publications, including in Nature (2), Science (1), Nature Genetics (3), Nature Neurosciences (1), PNAS (5), Molecular Psychiatry (6), American Journal of Psychiatry (7), JAMA Psychiatry (4), Biological Psychiatry (4), PLOS Genetics (1) and other high-ranking journals (average impact factor >10).
Screening and recruitment of patients
Working in conjunction with collaborating sites at the University of Southampton and Charité – Universitätsmedizin Berlin, the student will oversee the recruitment of patients from London’s NHS Trusts, with supporting recruitment from NHS PICs (NHS Trusts and GP practices). Participants will be identified by the following means:
- Consent for Contact (C4C)
- Clinical care teams
- Call for Participants adverts
Once consent for contact has been obtained where relevant, contact will be made in person, by phone, email, or letter. The student/research team will briefly explain the study and the information sheet will be sent by email or post. If, after receiving and reading the information sheet, the individual is interested, they will be screened by either the trained student/researcher on the team or a member of CRN staff using brief clinical measures or a brief online questionnaire. Suitability will depend on whether the individual is deemed to have either Major Depressive Disorder (BDI > 20), Alcohol Use Disorder (AUDIT > 20), or Psychosis (ICD-10 F20-F29).
The student will also be involved in the running of some of the institute assessments which comprise a) administering neuropsychological assessments, b) preparing participants for an MR scan and administering task assessments, c) conducting semi-structured clinical interview on substance use, d) administering other psychometric measures and scales, and e) using breathalyser tests, and facilitating blood and urine sample collection.
Identification of neurobehavioural clusters within and across diagnoses
The student will identify neurobehavioural clusters by measuring brain networks of BOLD response to tasks assessing reward processing, executive control, and social emotional processing, as well as resting state, and investigating their structural correlates. She will identify correlations of network configurations with behavioural symptoms in an overall sample of 600 patients (200 per diagnosis) and 150 matched controls, who will be characterised using precision phenotyping involving structural (white matter connectivity, grey matter volume/density) and functional (BOLD response to behavioural tasks, resting state) neuroimaging, cognitive (e.g. IQ, executive function), behavioural (related to respective symptom groups), and clinical assessments (e.g. Development and Wellbeing Assessment). The student will then measure correlations between network configurations and behavioural symptoms, within and across diagnoses. The resulting neurobehavioral clusters will form the building blocks of a clinical stratification that is based upon biological mechanisms and defined by the relative influence of environmental factors.
The student will then develop this methodology by introducing more sensitive first-level and second-level neuroimaging analysis methods to the analysis of large-scale datasets. These include (i) a novel combination of deep learning methods with canonical correlation procedures to establish precise relations between behavioural symptom clusters and deep network activity, (ii) application of multivariate analysis to omit spatial smoothing for increased precision of functional neuroimaging analyses [1,2], (iii) introduction to the analysis of large datasets of Neurite Orientation Dispersion and Density Imaging (NODDI), a more sensitive method to measure white matter connectivity [3].
Validation of findings in a large comparable cohort (IMAGEN)
The student will then validate the clusters identified in the clinical samples in the IMAGEN dataset (www.imagen-europe.com), a large longitudinal population-based imaging-genetics cohort (N=2090 at baseline/14 years). Considering that psychiatric disorders constitute extreme ends of one or several normally distributed quantitative behavioural traits [4], the student will test if network configurations at the extreme end of the IMAGEN spectrum correspond to those identified in the clinical sample by investigating whether the relationship between network activity and behavioural symptoms replicates in extreme (psychopathological) phenotypes in IMAGEN. At the same time, she will test specificity of the neurobehavioural clusters in a normative population-based sample. She will then conduct further validation analyses in several independent datasets with proven accessibility, QC, and comparability to the STRATIFY data.
Identification of longitudinal trajectories of dysfunctional neurobehavioural clusters
The student will proceed to establish individual longitudinal trajectories of network development throughout adolescence and young adulthood, leading to the development of reinforcement-related psychopathology. To carry out longitudinal analyses she will attain comparability of brain structure and function across time points using symmetric imaging registration (i.e. similar preprocessing of all images) implemented in the latest version of Statistical Parametric Mapping. The student will then measure the degree of change of functional networks and brain structure at baseline, at 19, and at 22 years. These measures will be carried out for trajectories resulting in networks associated with affective symptoms, alcohol use, and persistence of psychotic symptoms. To investigate the effect of environmental factors on the development of psychopathology, the student will analyse correlations between network trajectories and environmental factors. She will then apply the deep learning CCA method previously described to identify combinations of environmental terms that predict risk of the development of dysfunctional networks.
Supervisors: Prof Gunter Schumann and Dr. Erin Burke Quinlan
Entry requirements: Standard PhD entry requirements - Applicants should have a Bachelors degree with 2:1 honours (or Overseas equivalent). A 2:2 degree may be considered only where applicants also offer a Masters with Merit.
Award types and eligibility: Fully funded for three years, Home/EU tuitions fees (studentship not available to Overseas applicants), annual stipend and some research and travel costs.
Further information:
About the IoPPN (link to http://www.kcl.ac.uk/ioppn/about/index.aspx)
Studying at the IoPPN (link to http://www.kcl.ac.uk/ioppn/study/index.aspx)
About the Centre for Population Neuroscience and Precision Medicine (https://www.kcl.ac.uk/ioppn/research/centres/pons/about)
Studying at the SGDP Centre (https://www.kcl.ac.uk/ioppn/depts/sgdp-centre)
KCL Researcher Development Programme (http://www.kcl.ac.uk/study/pg/school/RDP/training-and-development/Researcher-Development-Programme-2014-15.pdf)
How to apply:
Applicants must complete and submit an online admissions application, via the admissions portal by midnight (23:59 GMT), Sunday 25th August 2019.
On the ‘Choosing a programme’ page, please select ‘Social Genetic & Developmental Psychiatry Research MPhil/PhD (Full Time)'.
In your application, you will be asked to include:
- Academic Transcripts – where applicable, academic transcripts must be submitted with the online admissions application
- Details of your qualifications (you will need to attach copies)
- Details of previous employment
- A personal statement describing your interests and why you wish to apply for this project. Please include this as an attachment rather than using the text box.
- Academic References – all admissions applications require two supporting references. If the applicant is relying on his/her referees to submit references directly to the College after he/she has submitted his/her admissions application, then the applicant must ensure that their chosen referees are made aware of the funding deadline.
In the Funding section, please tick box 5 and include the following reference: SGDP-GS-2019
Please note there is no need to complete the Research Proposal section in your application as the project has already been set.
You are welcome to email gunter.schumann@kcl.ac.uk (supervisor) for more information regarding the project and studentship.
If you have any queries regarding the application process, please contact Olivia Rees, Postgraduate Research Administrative Assistant.
References must be received by the deadline for the applicant to be eligible.
Only shortlisted applicants will be contacted.
Closing date: Sunday 25th August (23:59 GMT).
Interviews: W/c 5th September