The application of multiple omics technologies in biomedical cohorts has the potential to reveal patient-level disease characteristics and individualized response to treatment. However, the scale and heterogeneous nature of multi-modal data makes integration and inference a non-trivial task. We developed a deep-learning-based framework, multi-omics variational autoencoders (MOVE), to integrate such data and applied it to a cohort of 789 people with newly diagnosed type 2 diabetes with deep multi-omics phenotyping from the DIRECT consortium. Using in silico perturbations, we identified drug–omics associations across the multi-modal datasets for the 20 most prevalent drugs given to people with type 2 diabetes with substantially higher sensitivity than univariate statistical tests. From these, we among others, identified novel associations between metformin and the gut microbiota as well as opposite molecular responses for the two statins, simvastatin and atorvastatin. We used the associations to quantify drug–drug similarities, assess the degree of polypharmacy and conclude that drug effects are distributed across the multi-omics modalities.

The presentation and underlying pathophysiology of type 2 diabetes (T2D) is complex and heterogeneous. Recent studies attempted to stratify T2D into distinct subgroups using data-driven approaches, but their clinical utility may be limited if categorical representations of complex phenotypes are suboptimal.

Excess liver iron content is common and is linked to the risk of hepatic and extrahepatic diseases. We aimed to identify genetic variants influencing liver iron content and use genetics to understand its link to other traits and diseases. First, we performed a genome-wide association study (GWAS) in 8,289 individuals from UK Biobank, whose liver iron level had been quantified by magnetic resonance imaging, before validating our findings in an independent cohort (n = 1,513 from IMI DIRECT). Second, we used Mendelian randomisation to test the causal effects of 25 predominantly metabolic traits on liver iron content. Third, we tested phenome-wide associations between liver iron variants and 770 traits and disease outcomes. We identified 3 independent genetic variants (rs1800562 [C282Y] and rs1799945 [H63D] in HFE and rs855791 [V736A] in TMPRSS6) associated with liver iron content that reached the GWAS significance threshold (p <5 × 10-8). The 2 HFE variants account for ∼85% of all cases of hereditary haemochromatosis. Mendelian randomisation analysis provided evidence that higher central obesity plays a causal role in increased liver iron content. Phenome-wide association analysis demonstrated shared aetiopathogenic mechanisms for elevated liver iron, high blood pressure, cirrhosis, malignancies, neuropsychiatric and rheumatological conditions, while also highlighting inverse associations with anaemias, lipidaemias and ischaemic heart disease. Our study provides genetic evidence that mechanisms underlying higher liver iron content are likely systemic rather than organ specific, that higher central obesity is causally associated with higher liver iron, and that liver iron shares common aetiology with multiple metabolic and non-metabolic diseases. Excess liver iron content is common and is associated with liver diseases and metabolic diseases including diabetes, high blood pressure, and heart disease. We identified 3 genetic variants that are linked to an increased risk of developing higher liver iron content. We show that the same genetic variants are linked to higher risk of many diseases, but they may also be associated with some health advantages. Finally, we use genetic variants associated with waist-to-hip ratio as a tool to show that central obesity is causally associated with increased liver iron content.