This repo contains pre-trained Meta-matching models. If you want to train your own meta-matching model from scratch, please visit our CBIG repo.
He T, An L, Feng J, Bzdok D, Eickhoff SB, Yeo BTT. Meta-matching: a simple approach to translate predictive models from big to small data. BioRxiv, 2020.08.10.245373, under review.
There is significant interest in using brain imaging to predict phenotypes, such as cognitive performance or clinical outcomes. However, most prediction studies are underpowered. We propose a simple framework – meta-matching – to translate predictive models from large-scale datasets to new unseen non-brain-imaging phenotypes in small-scale studies. The key consideration is that a unique phenotype from a boutique study likely correlates with (but is not the same as) related phenotypes in some large-scale dataset. Meta-matching exploits these correlations to boost prediction in the boutique study. We apply meta-matching to predict non-brain-imaging phenotypes from resting-state functional connectivity. Using the UK Biobank (N=36,848) and HCP (N=1,019) datasets, we demonstrate that meta-matching can greatly boost the prediction of new phenotypes in small independent datasets in many scenarios. For example, translating a UK Biobank model to 100 HCP participants yields an 8-fold improvement in variance explained with an average absolute gain of 4.0% (min=-0.2%, max=16.0%) across 35 phenotypes.
Please check the detailed readme under each folder.
- It contains first release of the meta-matching model.
See our LICENSE file for license rights and limitations (MIT).
Please contact He Tong at [email protected], Lijun An at [email protected], Pansheng Chen at [email protected] and Thomas Yeo at [email protected].
Happy researching!
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent