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

IMI DIRECT Consortium; Rosa Lundbye Allesøe; Agnete Troen Lundgaard; Ricardo Hernández Medina; Alejandro Aguayo-Orozco; Joachim Johansen; Jakob Nybo Nissen; Caroline Brorsson; Gianluca Mazzoni; Lili Niu; Jorge Hernansanz Biel; Cristina Leal Rodríguez; Valentas Brasas; Henry Webel; Michael Eriksen Benros; Anders Gorm Pedersen; Piotr Jaroslaw Chmura; Ulrik Plesner Jacobsen; Andrea Mari; Robert Koivula; Anubha Mahajan; Ana Vinuela; Juan Fernandez Tajes; Sapna Sharma; Mark Haid; Mun-Gwan Hong; Petra B Musholt; Federico De Masi; Josef Vogt; Helle Krogh Pedersen; Valborg Gudmundsdottir; Angus Jones; Gwen Kennedy; Jimmy Bell; E Louise Thomas; Gary Frost; Henrik Thomsen; Elizaveta Hansen; Tue Haldor Hansen; Henrik Vestergaard; Mirthe Muilwijk; Marieke T Blom; Leen M 't Hart; Francois Pattou; Violeta Raverdy; Soren Brage; Tarja Kokkola; Alison Heggie; Donna McEvoy; Miranda Mourby; Ian Forgie; Ewan Pearson; Simon Rasmussen; Soren Brunak

doi:10.1038/s41587-022-01520-x

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

IMI DIRECT Consortium, Rosa Lundbye Allesøe, Agnete Troen Lundgaard, Ricardo Hernández Medina, Alejandro Aguayo-Orozco, Joachim Johansen, Jakob Nybo Nissen, Caroline Brorsson, Gianluca Mazzoni, Lili Niu, Jorge Hernansanz Biel, Cristina Leal Rodríguez, Valentas Brasas, Henry Webel, Michael Eriksen Benros, Anders Gorm Pedersen, Piotr Jaroslaw Chmura, Ulrik Plesner Jacobsen, Andrea Mari, Robert KoivulaAnubha Mahajan, Ana Vinuela, Juan Fernandez Tajes, Sapna Sharma, Mark Haid, Mun-Gwan Hong, Petra B Musholt, Federico De Masi, Josef Vogt, Helle Krogh Pedersen, Valborg Gudmundsdottir, Angus Jones, Gwen Kennedy, Jimmy Bell, E Louise Thomas, Gary Frost, Henrik Thomsen, Elizaveta Hansen, Tue Haldor Hansen, Henrik Vestergaard, Mirthe Muilwijk, Marieke T Blom, Leen M 't Hart, Francois Pattou, Violeta Raverdy, Soren Brage, Tarja Kokkola, Alison Heggie, Donna McEvoy, Miranda Mourby, Ian Forgie, Ewan Pearson, Simon Rasmussen (Lead / Corresponding author), Soren Brunak (Lead / Corresponding author)

Population Health and Genomics

Research output: Contribution to journal › Article › peer-review

35 Citations (Scopus)

69 Downloads (Pure)

Abstract

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.

Original language	English
Pages (from-to)	399-408
Number of pages	10
Journal	Nature Biotechnology
Volume	41
Early online date	2 Jan 2023
DOIs	https://doi.org/10.1038/s41587-022-01520-x
Publication status	Published - Mar 2023

Keywords

Data integration
Machine learning
Systems biology
Type 2 diabetes

ASJC Scopus subject areas

Applied Microbiology and Biotechnology
Bioengineering
Molecular Medicine
Biotechnology
Biomedical Engineering

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1038/s41587-022-01520-xLicence: CC BY

Final published versionFinal published version, 3.3 MBLicence: CC BY

Cite this

IMI DIRECT Consortium, Allesøe, R. L., Lundgaard, A. T., Hernández Medina, R., Aguayo-Orozco, A., Johansen, J., Nissen, J. N., Brorsson, C., Mazzoni, G., Niu, L., Biel, J. H., Rodríguez, C. L., Brasas, V., Webel, H., Benros, M. E., Pedersen, A. G., Chmura, P. J., Jacobsen, U. P., Mari, A., ... Brunak, S. (2023). Discovery of drug-omics associations in type 2 diabetes with generative deep-learning models. Nature Biotechnology, 41, 399-408. https://doi.org/10.1038/s41587-022-01520-x

@article{6592d2e3e2974de8824e4f26c8267cf5,

title = "Discovery of drug-omics associations in type 2 diabetes with generative deep-learning models",

abstract = "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.",

keywords = "Data integration, Machine learning, Systems biology, Type 2 diabetes",

author = "{IMI DIRECT Consortium} and Alles{\o}e, {Rosa Lundbye} and Lundgaard, {Agnete Troen} and {Hern{\'a}ndez Medina}, Ricardo and Alejandro Aguayo-Orozco and Joachim Johansen and Nissen, {Jakob Nybo} and Caroline Brorsson and Gianluca Mazzoni and Lili Niu and Biel, {Jorge Hernansanz} and Rodr{\'i}guez, {Cristina Leal} and Valentas Brasas and Henry Webel and Benros, {Michael Eriksen} and Pedersen, {Anders Gorm} and Chmura, {Piotr Jaroslaw} and Jacobsen, {Ulrik Plesner} and Andrea Mari and Robert Koivula and Anubha Mahajan and Ana Vinuela and Tajes, {Juan Fernandez} and Sapna Sharma and Mark Haid and Mun-Gwan Hong and Musholt, {Petra B} and {De Masi}, Federico and Josef Vogt and Pedersen, {Helle Krogh} and Valborg Gudmundsdottir and Angus Jones and Gwen Kennedy and Jimmy Bell and Thomas, {E Louise} and Gary Frost and Henrik Thomsen and Elizaveta Hansen and Hansen, {Tue Haldor} and Henrik Vestergaard and Mirthe Muilwijk and Blom, {Marieke T} and {'t Hart}, {Leen M} and Francois Pattou and Violeta Raverdy and Soren Brage and Tarja Kokkola and Alison Heggie and Donna McEvoy and Miranda Mourby and Ian Forgie and Ewan Pearson and Simon Rasmussen and Soren Brunak",

note = "Funding Information: The work leading to this publication has received support from the Innovative Medicines Initiative Joint Undertaking under grant agreement 115317 (DIRECT), resources of which are composed of financial contribution from the European Union{\textquoteright}s Seventh Framework Programme (FP7/2007-2013) and EFPIA companies{\textquoteright} in kind contribution. Information on the initiatives and activities of the IMI-DIRECT research consortium is available at https://directdiabetes.org. R.L.A., A.T.L, R.H.M., J.J., V.B., H.W., J.N.N., C.B., G.M., L.N., P.J.C., U.P.J, K.B., S.R., and S.B. were supported also by the Novo Nordisk Foundation (grants NNF14CC0001 and NNF17OC0027594). Additionally, R.H.M. was supported by Novo Nordisk Foundation (grant NNF20SA0035590). Copyright: {\textcopyright} 2023. The Author(s).",

year = "2023",

month = mar,

doi = "10.1038/s41587-022-01520-x",

language = "English",

volume = "41",

pages = "399--408",

journal = "Nature Biotechnology",

issn = "1087-0156",

publisher = "Nature Publishing Group",

}

IMI DIRECT Consortium, Allesøe, RL, Lundgaard, AT, Hernández Medina, R, Aguayo-Orozco, A, Johansen, J, Nissen, JN, Brorsson, C, Mazzoni, G, Niu, L, Biel, JH, Rodríguez, CL, Brasas, V, Webel, H, Benros, ME, Pedersen, AG, Chmura, PJ, Jacobsen, UP, Mari, A, Koivula, R, Mahajan, A, Vinuela, A, Tajes, JF, Sharma, S, Haid, M, Hong, M-G, Musholt, PB, De Masi, F, Vogt, J, Pedersen, HK, Gudmundsdottir, V, Jones, A, Kennedy, G, Bell, J, Thomas, EL, Frost, G, Thomsen, H, Hansen, E, Hansen, TH, Vestergaard, H, Muilwijk, M, Blom, MT, 't Hart, LM, Pattou, F, Raverdy, V, Brage, S, Kokkola, T, Heggie, A, McEvoy, D, Mourby, M, Forgie, I, Pearson, E, Rasmussen, S & Brunak, S 2023, 'Discovery of drug-omics associations in type 2 diabetes with generative deep-learning models', Nature Biotechnology, vol. 41, pp. 399-408. https://doi.org/10.1038/s41587-022-01520-x

TY - JOUR

T1 - Discovery of drug-omics associations in type 2 diabetes with generative deep-learning models

AU - IMI DIRECT Consortium

AU - Allesøe, Rosa Lundbye

AU - Lundgaard, Agnete Troen

AU - Hernández Medina, Ricardo

AU - Aguayo-Orozco, Alejandro

AU - Johansen, Joachim

AU - Nissen, Jakob Nybo

AU - Brorsson, Caroline

AU - Mazzoni, Gianluca

AU - Niu, Lili

AU - Biel, Jorge Hernansanz

AU - Rodríguez, Cristina Leal

AU - Brasas, Valentas

AU - Webel, Henry

AU - Benros, Michael Eriksen

AU - Pedersen, Anders Gorm

AU - Chmura, Piotr Jaroslaw

AU - Jacobsen, Ulrik Plesner

AU - Mari, Andrea

AU - Koivula, Robert

AU - Mahajan, Anubha

AU - Vinuela, Ana

AU - Tajes, Juan Fernandez

AU - Sharma, Sapna

AU - Haid, Mark

AU - Hong, Mun-Gwan

AU - Musholt, Petra B

AU - De Masi, Federico

AU - Vogt, Josef

AU - Pedersen, Helle Krogh

AU - Gudmundsdottir, Valborg

AU - Jones, Angus

AU - Kennedy, Gwen

AU - Bell, Jimmy

AU - Thomas, E Louise

AU - Frost, Gary

AU - Thomsen, Henrik

AU - Hansen, Elizaveta

AU - Hansen, Tue Haldor

AU - Vestergaard, Henrik

AU - Muilwijk, Mirthe

AU - Blom, Marieke T

AU - 't Hart, Leen M

AU - Pattou, Francois

AU - Raverdy, Violeta

AU - Brage, Soren

AU - Kokkola, Tarja

AU - Heggie, Alison

AU - McEvoy, Donna

AU - Mourby, Miranda

AU - Forgie, Ian

AU - Pearson, Ewan

AU - Rasmussen, Simon

AU - Brunak, Soren

N1 - Funding Information: The work leading to this publication has received support from the Innovative Medicines Initiative Joint Undertaking under grant agreement 115317 (DIRECT), resources of which are composed of financial contribution from the European Union’s Seventh Framework Programme (FP7/2007-2013) and EFPIA companies’ in kind contribution. Information on the initiatives and activities of the IMI-DIRECT research consortium is available at https://directdiabetes.org. R.L.A., A.T.L, R.H.M., J.J., V.B., H.W., J.N.N., C.B., G.M., L.N., P.J.C., U.P.J, K.B., S.R., and S.B. were supported also by the Novo Nordisk Foundation (grants NNF14CC0001 and NNF17OC0027594). Additionally, R.H.M. was supported by Novo Nordisk Foundation (grant NNF20SA0035590). Copyright: © 2023. The Author(s).

PY - 2023/3

Y1 - 2023/3

N2 - 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.

AB - 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.

KW - Data integration

KW - Machine learning

KW - Systems biology

KW - Type 2 diabetes

UR - http://www.scopus.com/inward/record.url?scp=85145508974&partnerID=8YFLogxK

U2 - 10.1038/s41587-022-01520-x

DO - 10.1038/s41587-022-01520-x

M3 - Article

C2 - 36593394

SN - 1087-0156

VL - 41

SP - 399

EP - 408

JO - Nature Biotechnology

JF - Nature Biotechnology

ER -

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

Abstract

Keywords

ASJC Scopus subject areas

UN SDGs

Access to Document

Other files and links

Fingerprint

DIRECT: Diabetes Research on Patient Stratification (joint with 25 other partners)

Author Correction: Discovery of drug-omics associations in type 2 diabetes with generative deep-learning models

Cite this

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

Abstract

Keywords

ASJC Scopus subject areas

UN SDGs

Access to Document

Other files and links

Fingerprint

Projects

DIRECT: Diabetes Research on Patient Stratification (joint with 25 other partners)

Research output

Author Correction: Discovery of drug-omics associations in type 2 diabetes with generative deep-learning models

Cite this