Login / Signup

Discovery of drug-omics associations in type 2 diabetes with generative deep-learning models.

Rosa Lundbye AllesøeAgnete Troen LundgaardRicardo Hernández MedinaAlejandro Aguayo-OrozcoJoachim JohansenJakob Nybo NissenCaroline BrorssonGianluca MazzoniLili NiuJorge Hernansanz BielValentas BrasasHenry WebelMichael Eriksen BenrosAnders Gorm PedersenPiotr Jaroslaw ChmuraUlrik Plesner JacobsenAndrea MariRobert Wilhelm KoivulaAnubha MahajanAna ViñuelaJuan Fernandez TajesSapna SharmaMark HaidMun-Gwan HongPetra B MusholtFederico De MasiJosef VogtHelle Krogh PedersenValborg GudmundsdottirAngus JonesGwen KennedyJimmy BellE Louise ThomasGary S FrostHenrik ThomsenElizaveta HansenTue Haldor HansenHenrik VestergaardMirthe MuilwijkMarieke T BlomLeen M 't HartFrancois PattouVioleta RaverdySoren BrageTarja KokkolaAlison HeggieDonna McEvoyMiranda MourbyJane KayeAndrew T HattersleyTimothy McDonaldMartin RidderstråleMark WalkerIan ForgieGiuseppe N GiordanoImre PavoHartmut RuettenOluf PedersenTorben HansenEmmanouil DermitzakisPaul W FranksJochen M SchwenkJerzy AdamskiMark I McCarthyEwan PearsonKarina BanasikSimon RasmussenSoren Brunaknull null
Published in: Nature biotechnology (2023)
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.
Keyphrases