Song, Qidong (2024) A computational investigation of polygenic index and individual differences in intervention effect. PhD thesis, Birkbeck, University of London.
![[img]](https://eprints.bbk.ac.uk/style/images/fileicons/text.png) |
Text
Song Q, final thesis for library.pdf - Full Version Restricted to Repository staff only until 25 September 2026. Download (13MB) |
Abstract
This thesis employs computational methods to explore the relationship between genetic variation, neurocomputation, and behavioural variability within the context of language development. Its ultimate goal is to assess the mechanistic viability of using polygenic indexes to predict optimal behavioural interventions within an approach called precision education. Behaviour genetics has now established that a combination of environmental and genetic measures can predict a large proportion of individual differences in language and cognition. Recent molecular genetic research has demonstrated that polygenic indexes (PGIs) derived from genetic variants in linkage disequilibrium (LD) with common single nucleotide polymorphisms (SNPs) can predict around a quarter of the variation in general cognitive ability. However, we have a limited understanding of the mechanisms underlying these influences, and the principal challenge is to build mechanistic models that link different levels of description. Building multiscale computational models of development, which incorporate links from genes to brain to behaviour, is one response. I use computational methods in this project to explore the mechanistic basis for individual differences in language development and intervention effects. In Chapter 3, using an existing neuroanatomically constrained model of language development I found pathway-specific and architecture-wide constraints on computational parameters had comparable effects on the model’s performance and developmental trajectory. In Chapter 4, I simulated two parallel populations of networks showing uneven cognitive development profiles within a multiscale framework. In Chapter 5, I generated polygenic indexes showing small but significant power in predicting developmental outcomes from each network’s artificial genome and subgroup membership when early behaviour data is available. Non-coding genes included in the PGIs had negative effects on the predictive power and the PGIs showed less power in the sconed population. In Chapter 6, a battery of behavioural interventions was simulated and applied to the two populations. Interventions showed larger short-term effects across measures and larger effects on typical developing individuals. In Chapter 7, I found interventions had a larger effect on individuals with mid-range PGIs and PGIs were only able to explain a small portion of the variance in the effectiveness and the size of the portion explained dropped over time. The PGIs were unable to pick the optimal intervention for every individual in our population.
Metadata
| Item Type: | Thesis |
|---|---|
| Copyright Holders: | The copyright of this thesis rests with the author, who asserts his/her right to be known as such according to the Copyright Designs and Patents Act 1988. No dealing with the thesis contrary to the copyright or moral rights of the author is permitted. |
| Depositing User: | Acquisitions And Metadata |
| Date Deposited: | 17 Oct 2024 10:05 |
| Last Modified: | 17 Oct 2024 16:00 |
| URI: | https://eprints.bbk.ac.uk/id/eprint/54412 |
| DOI: | https://doi.org/10.18743/PUB.00054412 |
Statistics
Additional statistics are available via IRStats2.
Archive Staff Only (login required)
