GoKawiilThe MULTI Consortium
The MULTI Consortium is an ongoing initiative to integrate and consolidate existing multi-organ and multi-omics data, including imaging, genetics, metabolomics and proteomics. Building on existing consortia and studies, MULTI aims to curate and harmonize the data to model human ageing and disease at scale across the lifespan. Refer to Supplementary Table 1 for comprehensive information, including the complete list of data analysed and their respective sample characteristics. The participants provided written informed consent to the corresponding studies. The MULTI Consortium is approved by the Institutional Review Board at Columbia University (AAAV6751).
UKBB
UKBB46 is a population-based research initiative comprising around 500,000 individuals from the United Kingdom between 2006 and 2010. Ethical approval for the UKBB study has been secured, and information about the ethics committee can be found online (https://www.ukbiobank.ac.uk/learn-more-about-uk-biobank/governance/ethics-advisory-committee). The main sleep data used in this study were sleep duration (field ID: 1160) based on a self-reported questionnaire collected from all 500,000 participants at the baseline. The 7 brain MRIBAGs were derived from multi-organ MRI data at the second visit, 11 ProtBAGs and 5 MetBAGs were derived using plasma proteomics and metabolomics at the baseline. Finally, we also included individual plasma proteins and metabolites in our ProWAS and MetWAS, as described below.
For the primary variable of interest in the UKBB, sleep duration (data-field: 1160) was assessed using an ACE touchscreen questionnaire asking “About how many hours sleep do you get in every 24 h? (please include naps).” The participants entered a numeric value, which underwent basic quality control: responses of less than 1 h or more than 23 h were rejected, and values below 3 h or above 12 h triggered a confirmation prompt. If participants clicked the ‘Help’ button, they were instructed that, if their sleep duration varied substantially, they should report the average number of hours slept in a 24-hour day over the past 4 weeks. For this variable, the value −1 indicates ‘Do not know’ and the value −3 indicates ‘Prefer not to answer’; these were excluded in the current work.
For the multi-organ IDPs, we used multi-organ MRI data from eight organ systems and tissues (category ID: 100003), including the brain, heart, liver, pancreas, spleen, adipose and kidney, as well as eye OCT features. The MUSE atlas-derived brain IDPs from the T1-weighted MRI10 were used for the brain MRIBAG generation. We also used neural networks to analyse the raw cardiac MRI images in our previous study and returned them to the UKBB to extract heart-specific IDPs (category ID: 157), which were used to derive the heart MRIBAG. For the other organs’ IDPs, we used the pre-derived features from the UKBB (category ID: 105). For the plasma proteomics data, we downloaded the original data (category ID: 1838), which were analysed and made available to the community by the UKBB Pharma Proteomics Project (UKB-PPP)47. The initial quality control procedures were described in the original work48; we conducted additional quality-check steps as outlined in the ‘Proteome-wide associations’ section. We also imputed missing normalized protein expression values and defined the organ-specific proteins using the HPA platform (https://www.proteinatlas.org/), as detailed in our previous work. For the plasma metabolomics data, we downloaded the original data (category ID: 220), which were analysed and made available to the community by Nightingale Health in the UKBB. Additional quality check analyses were performed as detailed in the ‘Metabolome-wide associations’ section.
FinnGen
The FinnGen49 study is a large-scale genomics initiative that has analysed over 500,000 Finnish biobank samples and correlated genetic variation with health data to understand disease mechanisms and predispositions. The project is a collaboration between research organizations and biobanks within Finland and international industry partners. For the benefit of research, FinnGen generously made its GWAS findings accessible to the wider scientific community (https://www.finngen.fi/en/access_results). This research used the publicly released GWAS summary statistics (version R9), which became available on 11 May 2022, after harmonization by the consortium. No individual data were used in the current study.
FinnGen published the R9 version of GWAS summary statistics via REGENIE software (v.2.2.4)50, covering 2,272 DEs, including 2,269 binary traits and 3 quantitative traits. The GWAS model included covariates such as age, sex, the initial 10 genetic principal components and the genotyping batch. Genotype imputation was referenced on the population-specific SISu v.4.0 panel. We included GWAS summary statistics for 521 FinnGen DEs in our analyses (Supplementary Table 7).
PGC
... continue reading