001019822 001__ 1019822
001019822 005__ 20240116205230.0
001019822 0247_ $$2doi$$a10.1101/2023.12.05.569848
001019822 0247_ $$2datacite_doi$$a10.34734/FZJ-2023-05653
001019822 037__ $$aFZJ-2023-05653
001019822 1001_ $$0P:(DE-HGF)0$$aChen, Pansheng$$b0
001019822 245__ $$aMultilayer meta-matching: translating phenotypic prediction models from multiple datasets to small data
001019822 260__ $$c2024
001019822 3367_ $$0PUB:(DE-HGF)25$$2PUB:(DE-HGF)$$aPreprint$$bpreprint$$mpreprint$$s1705409559_30837
001019822 3367_ $$2ORCID$$aWORKING_PAPER
001019822 3367_ $$028$$2EndNote$$aElectronic Article
001019822 3367_ $$2DRIVER$$apreprint
001019822 3367_ $$2BibTeX$$aARTICLE
001019822 3367_ $$2DataCite$$aOutput Types/Working Paper
001019822 520__ $$aResting-state functional connectivity (RSFC) is widely used to predict phenotypic traits in individuals. Large sample sizes can significantly improve prediction accuracies. However, for studies of certain clinical populations or focused neuroscience inquiries, small-scale datasets often remain a necessity. We have previously proposed a "meta-matching" approach to translate prediction models from large datasets to predict new phenotypes in small datasets. We demonstrated large improvement of meta-matching over classical kernel ridge regression (KRR) when translating models from a single source dataset (UK Biobank) to the Human Connectome Project Young Adults (HCP-YA) dataset. In the current study, we propose two meta-matching variants ("meta-matching with dataset stacking" and "multilayer meta-matching") to translate models from multiple source datasets across disparate sample sizes to predict new phenotypes in small target datasets. We evaluate both approaches by translating models trained from five source datasets (with sample sizes ranging from 862 participants to 36,834 participants) to predict phenotypes in the HCP-YA and HCP-Aging datasets. We find that multilayer meta-matching modestly outperforms meta-matching with dataset stacking. Both meta-matching variants perform better than the original "meta-matching with stacking" approach trained only on the UK Biobank. All meta-matching variants outperform classical KRR and transfer learning by a large margin. In fact, KRR is better than classical transfer learning when less than 50 participants are available for finetuning, suggesting the difficulty of classical transfer learning in the very small sample regime. The multilayer meta-matching model is publicly available at GITHUB_LINK.
001019822 536__ $$0G:(DE-HGF)POF4-5251$$a5251 - Multilevel Brain Organization and Variability (POF4-525)$$cPOF4-525$$fPOF IV$$x0
001019822 588__ $$aDataset connected to CrossRef
001019822 7001_ $$0P:(DE-HGF)0$$aAn, Lijun$$b1
001019822 7001_ $$0P:(DE-HGF)0$$aWulan, Naren$$b2
001019822 7001_ $$0P:(DE-HGF)0$$aZhang, Chen$$b3
001019822 7001_ $$0P:(DE-HGF)0$$aZhang, Shaoshi$$b4
001019822 7001_ $$0P:(DE-HGF)0$$aOoi, Leon Qi Rong$$b5
001019822 7001_ $$0P:(DE-HGF)0$$aKong, Ru$$b6
001019822 7001_ $$0P:(DE-HGF)0$$aChen, Jianzhong$$b7
001019822 7001_ $$0P:(DE-Juel1)177058$$aWu, Jianxiao$$b8
001019822 7001_ $$0P:(DE-HGF)0$$aChopra, Sidhant$$b9
001019822 7001_ $$0P:(DE-Juel1)136848$$aBzdok, Danilo$$b10
001019822 7001_ $$0P:(DE-Juel1)131678$$aEickhoff, Simon B$$b11
001019822 7001_ $$0P:(DE-HGF)0$$aHolmes, Avram J$$b12
001019822 7001_ $$0P:(DE-HGF)0$$aYeo, B. T. Thomas$$b13$$eCorresponding author
001019822 773__ $$a10.1101/2023.12.05.569848
001019822 8564_ $$uhttps://juser.fz-juelich.de/record/1019822/files/2023.12.05.569848v1.full.pdf$$yOpenAccess
001019822 8564_ $$uhttps://juser.fz-juelich.de/record/1019822/files/2023.12.05.569848v1.full.gif?subformat=icon$$xicon$$yOpenAccess
001019822 8564_ $$uhttps://juser.fz-juelich.de/record/1019822/files/2023.12.05.569848v1.full.jpg?subformat=icon-1440$$xicon-1440$$yOpenAccess
001019822 8564_ $$uhttps://juser.fz-juelich.de/record/1019822/files/2023.12.05.569848v1.full.jpg?subformat=icon-180$$xicon-180$$yOpenAccess
001019822 8564_ $$uhttps://juser.fz-juelich.de/record/1019822/files/2023.12.05.569848v1.full.jpg?subformat=icon-640$$xicon-640$$yOpenAccess
001019822 909CO $$ooai:juser.fz-juelich.de:1019822$$popenaire$$popen_access$$pVDB$$pdriver$$pdnbdelivery
001019822 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)177058$$aForschungszentrum Jülich$$b8$$kFZJ
001019822 9101_ $$0I:(DE-HGF)0$$6P:(DE-Juel1)177058$$a HHU Düsseldorf$$b8
001019822 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)131678$$aForschungszentrum Jülich$$b11$$kFZJ
001019822 9101_ $$0I:(DE-HGF)0$$6P:(DE-Juel1)131678$$a HHU Düsseldorf$$b11
001019822 9131_ $$0G:(DE-HGF)POF4-525$$1G:(DE-HGF)POF4-520$$2G:(DE-HGF)POF4-500$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$9G:(DE-HGF)POF4-5251$$aDE-HGF$$bKey Technologies$$lNatural, Artificial and Cognitive Information Processing$$vDecoding Brain Organization and Dysfunction$$x0
001019822 9141_ $$y2024
001019822 915__ $$0StatID:(DE-HGF)0510$$2StatID$$aOpenAccess
001019822 915__ $$0LIC:(DE-HGF)CCBYNCND4$$2HGFVOC$$aCreative Commons Attribution-NonCommercial-NoDerivs CC BY-NC-ND 4.0
001019822 920__ $$lyes
001019822 9201_ $$0I:(DE-Juel1)INM-7-20090406$$kINM-7$$lGehirn & Verhalten$$x0
001019822 980__ $$apreprint
001019822 980__ $$aVDB
001019822 980__ $$aUNRESTRICTED
001019822 980__ $$aI:(DE-Juel1)INM-7-20090406
001019822 9801_ $$aFullTexts