Research group of pharmacogenomics
Our research group aims to identify, characterize and quantify interindividual variability in drug response by comprehensive analysis of the human genome and health data. We combine data on drug plasma concentrations, treatment failure, adverse drug reactions, and genomic data from participants of the Estonian Biobank. We use text-mining tools to extract treatment outcomes from electronic health records, and have developed different methods for the analysis of longitudinal effects of medication use, and how genetic variants influence this variability.
The identified associations are tested using functional and pharmacokinetic studies. The results will serve as a basis for variants that should be included for preemptive pharmacogenomic testing, which could ultimately reduce the health and economic burden of low drug efficacy and unnecessary side effects caused by genetic variants.
Head of the Estonian Biobank, University of Tartu. Member of the Council of the Nordic Society of Human Genetics and Precision Medicine, Founding Member of the Estonian Academy for Young Researchers.
Contact: lili.milani@ut.ee CV
Postdoc completed at the University of Geneva (M. Dermitzakis Lab) from 2019 to 2022 on deciphering the genetic underpinnings of schizophrenia through multi-omics and deep phenotype data.
Contact: maris.alver@ut.ee CV
Research fellows
Silva Kasela, PhD
(Molecular Biomedicine)
Projects: SafePolyMed, PGxOMICS
Postdoc completed at New York Genome Center and Columbia University from 2018 to 2020 on genetic regulation of molecular phenotypes and their role in the development of complex diseases.
Burak Yelmen, PhD
(Genomics, Deep Learning)
Projects: REALMENT, PGxOMICS, MediMENT
Postdoc completed at LISN, Paris-Saclay University (2021-2023) on generating artificial human genomes and deep generative models in genomics.
Post Docs
PhD students
Tuuli Puusepp, MSc
(Mathematical Statistics)
Hanna Maria Kariis, MSc
(Neuroscience)
Supervisors: Lili Milani, Kelli Lehto, Maris Alver
Projects: Realment
Saskia Kuusk, MSc
(Mathematical Statistics)
Supervisors: Krista Fischer, Lili Milani
Projects: Cholesterol-lowering therapy and cardiovascular diseases
Other staff
Project manager, assistant to Lili Milani
All projects of the group, administrative issues
Visiting Researchers
|
Image
| Benjamin Crapo Stone (2023–2024) Fulbright scholarship BSc student from Brigham Young University-Provo, USA Project: Predicting the relation between cardiovascular disease and mental health conditions Supervisors: Burak Yelmen, Lili Milani | |
|
Image
| Robin Hofmeister, PhD Computational Biology (2024- 2026) Projects: Using haplotype data to improve complex traits and diseases analysis PhD supervised by Olivier Delaneau at the University of Lausanne, Switzerland, from 2019 to 2023 on inferring the parental origin of haplotypes in biobanks. Current employment: University Center for Primary care and Public Health, Lausanne, supervised by Zoltán Kutalik and as a visiting researcher at University of Tartu, supervised by Lili Milani. Contact: robin.hofmeister@unil.ch, CV |
Hanna Maria Kariis “Genetic studies of the comorbidity between cardiovascular disease and mental disorders: from mechanisms to improved treatment”
Supervisors: Lili Milani; Kelli Lehto; Tõnis Org; Maris Alver
Siim Kurvits “Predictive modelling of cardiometabolic diseases in psychiatric patients: towards improved prevention using individual-level genetic data”
Supervisors: Kelli Lehto; Lili Milani; Toomas Haller
Tuuli Puusepp “Models for risk prediction in complex data structures”
Supervisors: Krista Fischer; Lili Milani; Meelis Käärik
Ida Maria Orula “Haplotype-based pharmacogenomic studies based on electronic health records and genetic data”
Supervisors: Sulev Reisberg; Lili Milani
Martin Meitern “Using Genetic Information as an Input in Changing Health Behaviour”
Supervisors: Sten Hansson; Lili Milani; Andero Uusberg
Ave Põld “Personalised prevention of ischemic heart disease among high polygenic risk individuals”
Supervisors: Mikk Jürisson, Lili Milani, Krista Fischer
Jelisaveta Džigurski “Pharmacogenetic aspects of hormonal contraceptives”
Supervisors: Triin Laisk; Lili Milani; Reedik Mägi
Laura Birgit Luitva “Models for drug-drug interactions, population pharmacokinetics and pharmacogenetics”
Supervisors: Krista Fischer; Maris Alver; Lili Milani
- Merve Nur Güler “Deep learning for predicting quantitative traits from genomic data”
Supervisors: Burak Yelmen; Lili Milani; Luca Pagani
- Saskia Kuusk “Causal modelling in biobank cohorts: methodological issues and applications”
Supervisors: Krista Fischer, Lili Milani
- Kristi Krebs (PhD, 2020)
Exploring the genetics of adverse events in pharmacotherapy using Biobanks and Electronic Health Records
Supervisors: Lili Milani; Andres Metspalu - Epp Kaleviste (PhD, 2020)
Genetic variants revealing the role of STAT1/STAT3 signaling cytokines in immune protection and pathology
Supervisors: Kai Kisand; Pärt Peterson; Lili Milani - Tõnis Tasa (PhD, 2019)
Bioinformatics Approaches in Personalised Pharmacotherapy
Supervisors: Jaak Vilo; Lili Milani; Tuuli Metsvaht - Silva Kasela (PhD, 2017)
Genetic regulation of gene expression: detection of tissue- and cell type-specific effects
Supervisors: Lili Milani; Krista Fischer; Andres Metspalu - Riin Tamm (PhD, 2017)
In-depth analysis of factors affecting variability in thiopurine methyltransferase activity
Supervisors: Andres Metspalu; Lili Milani - Kaie Lokk (PhD, 2017)
Comparative genome-wide DNA methylation studies of healthy human tissues and non-small cell lung cancer tissue
Supervisors: Neeme Tõnisson; Lili Milani - Saskia Kuusk (MA, 2024)
Target trial emulation: the effect of polygenic risk score and cholesterol-lowering therapy on cardiovascular disease risk
Supervisors: Krista Fischer, Lili Milani - Laura Birgit Luitva (MA, 2022)
Relationship between Depression and Cardiometabolic Diseases Using Multi-State Models Based on Estonian Genome Centre's Data
Supervisors: Silva Kasela; Lili Milani - Elisabeth Kelner (MA, 2022)
The impact of polygenic risk score on the severity and the use of medication on type two diabetes
Supervisors: Mikk Jürisson; Heti Pisarev; Vallo Volke; Lili Milani - Kadri Maal (MA, 2021)
Cytochrome P450 2C19 deletions in the Estonian population
Supervisors: Lili Milani; Tarmo Puurand - Liis Haljasmägi (MA, 2018)
Type I IFN neutralizing autoantibodies in patients with systemic lupus erythematosus
Supervisors: Kai Kisand; Lili Milani; Pärt Peterson - Marili Palover (MA, 2016)
Associations between telomere length and DNA methylation levels
Supervisors: Lili Milani; Silva Kasela - Hanna Rein (MA, 2016)
TaqMan technology based methylation pattern analysis to monitore premature immune system aging
Supervisors: Kai Kisand; Lili Milani - Kelli Grand (MA, 2016)
Analysing loss-of-function mutations by pairing 2300 whole genomes with electronic health records
Supervisors: Lili Milani; Pärt Peterson - Epp Kaleviste (MA, 2015)
The impact of STAT1 gain-of-function mutation on JAK-STAT signaling pathway
Supervisors: Lili Milani; Kai Kisand - Liina Pappa (MA, 2014)
Exome sequencing of two patients with unilateral renal hypoplasia
Supervisors: Lili Milani; Tiit Nikopensius - Mari-Liis Reim (MA, 2013)
Factors influencing the activity of thiopurine methyltransferase in human liver
Supervisors: Riin Tamm; Lili Milani; Andres Metspalu - Silva Kasela (MA; 2013)
DNA methylation: normalisation and analysis
Supervisors: Krista Fischer; Lili Milani
Overview of ongoing projects
Using real-world big data from eHealth, biobanks and national registries, integrated with clinical trial data to improve outcome of severe mental disorders (1.06.2021−30.11.2025); Principal Investigator: Lili Milani; Horizon 2020 project
Financier: Commission of the European Communities; Financing: 444 000 EUR. (Project coordinator: Ole Andreassen, University of Oslo, total budget 6M EUR)
The REALMENT project aims to develop innovative tools to individualize treatments using available psychiatric medication. Big data from populations, cohorts and eHealth samples and whole genome genotypes will be analyzed in an EU-wide sustainable infrastructure. Artificial intelligence and machine learning methods will be used to develop prediction and stratification tools for precision psychiatry. For implementation, the clinical management platform (4MENT) will be made available to provide decision support to clinicians for optimizing therapeutic effects.
In addition to the CoMorMent consortium members, REALMENT project includes enterprises that can facilitate the implementation of project results to practice.
Data for analysis: Big data from populations (Nordic registries), cohorts (European biobanks), and eHealth samples (medical records), whole genome genotypes (n=1.8 million). Validation will be performed in large RCT data (n=10k) cohorts.
Project website: https://www.realment.uio.no/
Improve Safety in Polymedication by Managing Drug-Drug-Gene Interactions (1.06.2022−30.11.2025); Principal Investigator: Lili Milani; Horizon Europe project
Financier: Commission of the European Communities; Financing: 598 000 EUR. (Project coordinator: Thorsten Lehr, University of Saarland, total budget 5,6 M EUR)
The project will develop a framework to define, assess and manage real-life drug-drug-gene interactions for physicians and individual patients. The main objectives are to use machine learning methods to develop an evidence-based risk scoring system, an electronic tool for patients to manage their therapies, create a mathematical model for effective dose adaptions for clinically relevant compounds and validate it. The vast consortium is largely based in Germany, but includes collaborators from Greece, Estonia and Finland.
Data for analysis: genotypes, linking to electronic health records (EHRs), self-reported questionnaires’ (health, environmental/life-style) data
Project website: https://www.safepolymed.eu/
Preventing non-communicable diseases caused by the post-acute phase of cOvid-19 INfecTion (1.01.2024−31.12.2027); Principal Investigator: Pärt Peterson, Senior Researcher: Lili Milani; Horizon Europe project
Financier: Commission of the European Communities; Financing: 96 000 EUR (Project coordinator: Claus Desler Madsen, University of Copenhagen, total budget 9,0 M EUR, total budget for University of Tartu 771 250 EUR).
The project will develop knowledge-based biomarkers for the prevention and management of non-communicable diseases with a virtual twin model that offers clinical decision support, and clinical guidelines and recommendations for healthcare. The goal is to diminish the effects of the COVID-19 post-acute phase.
Data for analysis: genotyped biobank samples united with large national registries with information about disease trajectories and comorbidity, more than 6 million Europeans, and cross-sectional biobanks from more than 6000 Europeans.
Project website: https://cordis.europa.eu/project/id/101137196
Centre for Data Enriched Medicine (TeamPerMed) (1.09.2023−31.08.2029); Principal Investigator: Mait Metspalu, Senior Researcher: Lili Milani; HorizonEurope project
Financier: Commission of the European Communities, Estonian Ministry of Education and Research; Financing: 200 000 EUR (Total budget 6,3 M EUR, thereof 3,3 M EUR from EC).
Centre for Personalised Medicine (TeamPerMed) will be a multi-disciplinary centre that integrates expertise in Genomics, IT, Clinical Medicine and Socio-Economic Analysis to create a scalable framework for developing clinical guidelines and clinical decision support tools that can be effectively integrated into healthcare systems.
Lili Milani will coordinate the clinical trial of pharmacogenomics.
Partners: Erasmus Medical Centre, the Netherlands, and the University of Helsinki, Finland.
Data for analysis: genotyped biobank samples united with large national registries with information about disease trajectories and comorbidity.
Project website: will be updated.
Computational drug repurposing based on EHR and large Biobanks (MediMENT) (1.01.2025-31.12.2027); Principal Investigator: Lili Milani, Senior Researcher: Burak Yelmen; ERA-NET project.
Financier: Estonian Research Council and European Partnership for Personalized Medicine; Financing: 150 000 EUR (Total budget 1,4 M EUR).
Mental disorders are among the largest chronic disease groups worldwide, with a great need for more knowledge to improve treatment and care. There is a shortage of drugs to treat mental disorders due to many reasons, including the lengthy and high-risk process involved in drugs' development. A way to accelerate drug discovery is drug repurposing. This strategy involves uncovering new pharmacological effects for already approved drugs, offering the potential to streamline drug development by bypassing certain stages. Computational drug repurposing uses sources like omics data (e.g., gene and protein expression), biomedical association knowledgebase and electronic health records. In this project, we aim to discover drugs that could potentially be used to treat mental disorders. Project partners are from Denmark, Norway, Sweden, Iceland and Estonia.
Data for analysis: genotyped biobank samples united with large national registries with information about disease trajectories and comorbidity.
Project website: will be updated.
Population-specific Prediction Models for Metabolic Syndrome and Treatment Response for Schizophrenia Spectrum Disorder Patients Using Clinical Data and Electronic Health Records (SCZMetS) (1.01.2025−31.12.2029); Principal Investigator: Maris Alver; national research project
Financier: Estonian Research Council; Financing: 585 000 EUR.
Schizophrenia spectrum disorder (SSD) patients exhibit a higher prevalence of metabolic syndrome (MetS) compared to the general population. Despite extensive research, clinical assessment of MetS in SSD patients remains inadequate, resulting in suboptimal risk estimation and treatment discontinuation. Our goal is to address this gap by developing population-specific prediction models for MetS and treatment response tailored for SSD patients. By leveraging thorough assessment of longitudinal clinical data, utilizing large-scale electronic health records, integrating genetics, and applying external validation, we aim to enhance risk assessment and management strategies of MetS and antipsychotic treatment response in SSD patients, ultimately improving health outcomes in this vulnerable population and providing further clarity of the molecular mechanisms underlying this comorbidity.
Data for analysis: Estonian biobank and UK biobank genotype and electronic health records, additional data gathered during medical check-up.
Next-generation pharmacogenomics: systematic integration of genetics, physiology, and drug-drug interactions (PGxOMICS) (1.01.2025-31.12.2029); Principal Investigator: Lili Milani; national research project
Financier: Estonian Research Council; Financing: 1 346 500 EUR.
We plan to integrate data on systematically assessed genetic variation, physiology, co-medications, age, and sex to explore their interactions and impact on treatment outcomes using real-world data of large-scale biobanks coupled with comprehensively annotated genomic information. Leveraging extensive phenotype data mined from electronic health records using large language models, alongside the discovery, validation and improved imputation of novel and known genetic variants will allow systematic modelling across genetic and physiological domains. This approach lays the groundwork for developing advanced risk prediction models for pharmacogenomics, ultimately leading to improved personalized treatment for individuals.
Data for analysis: Estonian biobank and UK biobank genotype and electronic health records
Overview of recently completed projects
Predicting comorbid cardiovascular disease in individuals with mental disorder by decoding disease mechanisms (1.01.2020−31.12.2024); Principal Investigator: Lili Milani; Horizon 2020 project
Financier: Commission of the European Communities; Financing: 554 000 EUR. (Project coordinator: Ole Andreassen, University of Oslo, total budget 5,9 M EUR)
The Nordics-centered project is designed to uncover mechanisms that cause higher incidence of cardiovascular disease in people with mental disorders. It considers genetic and lifestyle risks and their impact on mental disorders. During the project, structural brain changes and body fat composition from MRI data will be analyzed in combination with gene expression and functional studies. CoMorMent has a multidisciplinary expert team in clinical science, genetic epidemiology, molecular genetics, and neuroscience combined with experts in machine learning and computation.
Data for analysis: genotyped biobank samples united with large national registries with information about disease trajectories and comorbidity in over 1.8 million people belonging to the biobanks of the Nordics, UK and Estonia.
Project website: https://www.comorment.uio.no/
(1.01.2022−31.12.2024); Principal Investigator: Lili Milani; Swedish Research Council funded project
Financier: Swedish Research Council; Financing: 57 000 EUR (Project coordinator: Magnus Ingelman-Sundberg, Karolinska Institutet, total budget 3,6 M SEK)
The project aims to identify, characterize and quantify unexplained genetically-predicted interindividual variability in drug response which will enable clinicians to perform precise dosing for each individual patient. Three labs from Sweden, Norway and Estonia will analyze the data on plasma concentrations, treatment failure, adverse drug reactions, and genetic sequences. Special emphasis will be placed on the identification of novel haplotypes, globally identified mutations of importance for adverse drug reactions response, the role of rare mutations and the development of a novel genetic platform. The project aims to deliver a novel diagnostic tool, in compliance with the FAIR principles and GDPR, to be implemented for decision support by integrating AI-derived pharmacogenomics-based algorithms into existing electronic infrastructure.
Data for analysis: host genotypes, linking to electronic health records (EHRs), self-reported questionnaires’ (health, environmental/life-style) data from > 20,000 patients.
(1.01.2018−30.06.2023); Principal Investigator: Lili Milani; national research project
Financier: Estonian Research Council; Financing: 835 000 EUR.
Interindividual variability in drug metabolism and sensitivity for drug toxicity persists as a major problem for drug treatment. Recent research has highlighted the large extent of rare variants in genes with importance for drug metabolism. The aim of this study is to create a catalog of common and rare genetic variants related to suboptimal drug metabolism and adverse drug reactions (ADRs) in the Estonian and Swedish populations. This will be achieved using genome sequences of 5000 and genotype data of 50,000 individuals combined with extensive health records regarding drug prescriptions and ADR diagnoses. The identified gene-ADR associations will be tested using functional studies. The results will serve as a basis for variants that can be included for preemptive pharmacogenomic testing, which could ultimately reduce the health and economic burden of low drug efficacy and unnecessary side effects caused by genetic variants.
Data for analysis: Estonian biobank and UK biobank genotype and electronic health records, additional data on drug metabolism pharmacokinetics gathered during a clinical trial.
(1.10.2021−30.06.2023); Project Investigator: Lili Milani; national research project
Financier: Ministry of Education and Research; Financing: 164 000 EUR (Project coordinator: Marje Oona, University of Tartu, total budget 555 000 EUR).
The objective of the study is to assess the need for booster doses of COVID-19 vaccines. The specific aims are: to assess the persistence of immunity after COVID-19 vaccination and the factors influencing it among the adult population in Estonia, and to assess the persistence of immunity after SARS-CoV-2 infection and the factors influencing it in the Estonian adult population.
Personalised Medicine by Predictive Modeling in Stroke for better Quality of Life (1.05.2018−30.04.2023); Principal Investigator: Lili Milani; Horizon 2020 project
Financier: Commission of the European Communities; Financing: 366 000 EUR. (Project coordinator: Dietmar Frey, CHARITE Berlin, total budget 6,0 M EUR)
The project integrated various data from different sources: genomics, microbiomics, biochemical; imaging, social, lifestyle, gender; economic and worklife.
Novel decision support models to address the needs of the patient in all stages of the disease (1. Prevention, 2. Acute treatment, 3. Rehabilitation, 4. Reintegration) were created. PRECISE4Q was designed to have a clinically measurable and sustainable impact, leading to better understanding of risk, health and resilience factors and support patients throughout their life-long journey by personalised strategies for their specific needs.
Data for analysis: genotyped biobank samples united with large national registries with information about disease trajectories and comorbidity.
Project website: https://precise4q.eu/
Funding
Commission of the European Communities:
- 366 000 EUR, H2020 project PRECISE4Q (1.05.2018−30.04.2023)
- 444 000 EUR, H2020 project CoMorMent (1.01.2020−30.06.2024)
- 554 000 EUR, H2020 project Realment (1.06.2021−31.05.2025)
- 598 000 EUR, Horizon Europe project SafePolyMed (1.06.2022−30.11.2025)
- 96 000 EUR, Horizon Europe project POINT (1.01.2024−31.12.2027)
- 200 000 EUR, Horizon Europe project TeamPerMed (1.09.2023−31.08.2029)
Swedish Research Council:
- 57 000 EUR, Genetic Precision Medicine project (1.01.2022−31.12.2024)
Estonian Research Council and European Partnership for Personalized Medicine (ERA_NET):
- 150 000 EUR, ERA-NET project MediMENT (1.01.2025-31.12.2027)
Estonian Research Council:
- 835 000 EUR, Role of Rare and Common Genetic Variants in Adverse Drug Reactions project (1.01.2018−30.06.2023)
- 585 000 EUR, Population-specific Prediction Models for Metabolic Syndrome and Treatment Response for Schizophrenia Spectrum Disorder Patients Using Clinical Data and Electronic Health Records (SCZMetS) (1.01.2025−31.12.2029); Principal Investigator: Maris Alver
- 1 346 500 EUR, .Next-generation pharmacogenomics: systematic integration of genetics, physiology, and drug-drug interactions (PGxOMICS) (1.01.2025-31.12.2029);
Ministry of Education and Research:
- 164 000 EUR, COVID-19 longitudinal study (1.10.2021−30.06.2023)
Major publications
Alver M, Kasela S, Haring L, Luitva LB; Estonian Biobank Research Team; Health Informatics Research Team; Fischer K, Möls M, Milani L. Genetic predisposition and antipsychotic treatment effect on metabolic syndrome in schizophrenia: a ten-year follow-up study using the Estonian Biobank. Lancet Reg Health Eur. 2024 Apr 26;41:100914. doi: 10.1016/j.lanepe.2024.100914. PMID: 38707868.
Kiiskinen T, Helkkula P, Krebs K, Karjalainen J, Saarentaus E, Mars N, Lehisto A, Zhou W, Cordioli M, Jukarainen S, Rämö JT, Mehtonen J, Veerapen K, Räsänen M, Ruotsalainen S, Maasha M; FinnGen; Niiranen T, Tuomi T, Salomaa V, Kurki M, Pirinen M, Palotie A, Daly M, Ganna A, Havulinna AS, Milani L, Ripatti S. Genetic predictors of lifelong medication-use patterns in cardiometabolic diseases. Nat Med. 2023 Jan;29(1):209-218. doi: 10.1038/s41591-022-02122-5.
Shen Q, Mikkelsen DH, Luitva LB, Song H, Kasela S, Aspelund T, Bergstedt J, Lu Y, Sullivan PF, Ye W, Fall K, Tornvall P, Pawitan Y, Andreassen OA, Buil A, Milani L, Fang F, Valdimarsdóttir U. Psychiatric disorders and subsequent risk of cardiovascular disease: a longitudinal matched cohort study across three countries. EClinicalMedicine. 2023 Jun 22;61:102063. doi: 10.1016/j.eclinm.2023.102063.
Kariis HM, Kasela S, Jürgenson T, Saar A, Lass J, Krebs K, Võsa U, Haan E; Estonian Biobank Research Team. The role of depression and antidepressant treatment in antihypertensive medication adherence and persistence: Utilising electronic health record data. J Psychiatr Res. 2023 Oct 27;168:269-278. doi: 10.1016/j.jpsychires.2023.10.018.
Mishra, A., Malik, R., Hachiya, T, Jürgenson T. et al. Stroke genetics informs drug discovery and risk prediction across ancestries. Nature 611, 115–123 (2022). https://doi.org/10.1038/s41586-022-05165-3
Krebs K, Bovijn J, Zheng N, et al. Genome-wide Study Identifies Association between HLA-B∗55:01 and Self-Reported Penicillin Allergy. Am J Hum Genet. 2020;107(4):612-621. doi:10.1016/j.ajhg.2020.08.008
Tasa T, Krebs K, Kals M, et al. Genetic variation in the Estonian population: pharmacogenomics study of adverse drug effects using electronic health records. Eur J Hum Genet. 2019;27(3):442-454. doi:10.1038/s41431-018-0300-6
Reisberg S, Krebs K, Lepamets M, et al. Translating genotype data of 44,000 biobank participants into clinical pharmacogenetic recommendations: challenges and solutions. Genet Med. 2019;21(6):1345-1354. doi:10.1038/s41436-018-0337-5
Kasela S, Kisand K, Tserel L, et al. Pathogenic implications for autoimmune mechanisms derived by comparative eQTL analysis of CD4+ versus CD8+ T cells. PLoS Genet. 2017;13(3):e1006643. Published 2017 Mar 1. doi:10.1371/journal.pgen.1006643
Participation in consortia and working groups
- Psychiatric Genomics Consortium (Milani)
- eQTLGen Consortium (Milani, Kasela)
- Genetics of DNA Methylation (GoDMC) Consortium (Milani, Kasela)
- NHLBI Trans-Omics for Precision Medicine (TOPMed) (Kasela)
- Genotype-Tissue Expression (GTEx) Consortium (Kasela)
- American Society of Human Genetics (Alver, Kasela, Milani)
- European Society of Human Genetics (Alver, Milani)
- Estonian Society of Human genetics (Alver, Krebs)
- Estonian Statistical Society (Kasela)
