Search

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. Clinical multi-omics data are integrated and analyzed using a generative deep-learning model. [ABSTRACT FROM AUTHOR]

Introduction: It is uncertain whether lixisenatide has postprandial insulinotropic effects when its effect on slowing gastric emptying is considered, in healthy subjects and type 2 diabetes mellitus (T2DM). We evaluated the effects of single administration of 10 μg sc lixisenatide on glycaemia, insulin secretion and gastric emptying (GE), measured using the 'gold standard' technique of scintigraphy following an oral glucose load (75 g glucose). Methods: Fifteen healthy subjects (nine men, six women; age 67.2 ± 2.3 years) and 15 patients with T2DM (nine men, six women; age 61.9 ± 2.3 years) had measurements of GE, plasma glucose, insulin and C-peptide for 180 min after a radiolabeled 75 g glucose drink on two separate days. All subjects received lixisenatide (10 μg sc) or placebo in a randomised, double-blind, crossover fashion 30 min before the drink. Insulin secretory response (ISR) was determined using the C-peptide deconvolution method. Results: GE was markedly slowed by lixisenatide compared with placebo in both healthy subjects (1.45 ± 0.10 kcal/min for placebo vs. 0.60 ± 0.14 kcal/min for lixisenatide) and diabetes (1.57 ± 0.06 kcal/min for placebo vs. 0.75 ± 0.13 kcal/min for lixisenatide) (both P < 0.001) with no difference between the two groups (P = 0.42). There was a moderate to strong inverse correlation between the early insulin secretory response calculated at 60 min and gastric retention at 60 min with lixisenatide treatment in healthy subjects (r = − 0.8, P = 0.0003) and a trend in type 2 diabetes (r = − 0.4, P = NS), compared with no relationships in the placebo arms (r = − 0.02, P = NS, healthy subjects) and (r = − 0.16, P = NS, type 2 diabetes). Conclusion: The marked slowing of GE of glucose induced by lixisenatide is associated with attenuation in the rise of postprandial glucose in both healthy subjects and diabetes and early insulin secretory response in healthy subjects. Clinical Trials Registration Number: NCT02308254. [ABSTRACT FROM AUTHOR]