A5053426515- Metabolomics and Mass Spectrometry Studies
- Genetic Associations and Epidemiology
- Bioinformatics and Genomic Networks
- Computational Drug Discovery Methods
- Liver Disease Diagnosis and Treatment
- COVID-19 Clinical Research Studies
- Cardiovascular Health and Risk Factors
- Tensor decomposition and applications
- Cell Image Analysis Techniques
- Computational Physics and Python Applications
- Machine Learning in Healthcare
- Gene expression and cancer classification
- Diabetes, Cardiovascular Risks, and Lipoproteins
- Long-Term Effects of COVID-19
- Obesity, Physical Activity, Diet
- Advanced Neural Network Applications
- Advanced Causal Inference Techniques
- Protein Structure and Dynamics
- Diet and metabolism studies
- Drug-Induced Hepatotoxicity and Protection
- Biomedical Text Mining and Ontologies
- Cancer, Lipids, and Metabolism
- SARS-CoV-2 and COVID-19 Research
- Machine Learning in Bioinformatics
- Genetics, Bioinformatics, and Biomedical Research
Aalto University
2018-2024
Helsinki Institute for Information Technology
2018-2020
Abstract Blood lipids and metabolites are markers of current health future disease risk. Here, we describe plasma nuclear magnetic resonance (NMR) biomarker data for 118,461 participants in the UK Biobank. The biomarkers cover 249 measures lipoprotein lipids, fatty acids, small molecules such as amino ketones, glycolysis metabolites. We provide an atlas associations these to prevalence, incidence, mortality over 700 common diseases ( nightingalehealth.com/atlas ). results reveal a plethora...
Biomarkers of low-grade inflammation have been associated with susceptibility to a severe infectious disease course, even when measured prior onset. We investigated whether metabolic biomarkers by nuclear magnetic resonance (NMR) spectroscopy could be pneumonia (2507 hospitalised or fatal cases) and COVID-19 (652 in 105,146 generally healthy individuals from UK Biobank, blood samples collected 2007-2010. The overall signature biomarker associations was similar for the risk COVID-19. A...
Abstract We present comboFM, a machine learning framework for predicting the responses of drug combinations in pre-clinical studies, such as those based on cell lines or patient-derived cells. comboFM models context-specific interactions through higher-order tensors, and efficiently learns latent factors tensor using powerful factorization machines. The approach enables to leverage information from previous experiments performed similar drugs cells when new so far untested cells; thereby, it...
Many inference problems in bioinformatics, including drug bioactivity prediction, can be formulated as pairwise learning problems, which one is interested making predictions for pairs of objects, e.g. drugs and their targets. Kernel-based approaches have emerged powerful tools solving that kind, especially multiple kernel (MKL) offers promising benefits it enables integrating various types complex biomedical information sources the form kernels, along with importance prediction task....
Pleiotropy and genetic correlation are widespread features in genome-wide association studies (GWAS), but they often difficult to interpret at the molecular level. Here, we perform GWAS of 16 metabolites clustered intersection amino acid catabolism, glycolysis, ketone body metabolism a subset UK Biobank. We utilize well-documented biochemistry jointly impacting these analyze pleiotropic effects context their pathways. Among 213 lead hits, find strong enrichment for genes encoding...
Identifying individuals at high risk of chronic diseases via easily measured biomarkers could enhance efforts to prevent avoidable illness and death. Using 'omic data can stratify for many simultaneously from a single measurement that captures multiple molecular predictors risk. Here we present nuclear magnetic resonance metabolomics in blood samples 700,217 participants three national biobanks. We built metabolomic scores identify high-risk groups cause the most morbidity high-income...
Abstract Identifying individuals at high risk of chronic diseases via easily measured biomarkers could improve public health efforts to prevent avoidable illness and death. Here we present nuclear magnetic resonance blood metabolomics from half a million samples three national biobanks. We built metabolomic scores that identify high-risk group for each 12 cause the most morbidity in high-income countries show consistent cross-biobank replication relative disease these groups. are more...
Abstract Blood lipids and metabolites are both markers of current health indicators risk for future disease. Here, we describe plasma nuclear magnetic resonance (NMR) biomarker data 118,461 participants in the UK Biobank, an open resource public research with extensive clinical genomic data. The biomarkers cover 249 measures lipoprotein lipids, fatty acids, small molecules such as amino ketones, glycolysis metabolites. We provide a systematic atlas associations these to prevalence,...
ABSTRACT Background Identification of healthy people at high risk for severe COVID-19 is a global health priority. We investigated whether blood biomarkers measured by high-throughput metabolomics could be predictive pneumonia and hospitalisation years after the sampling. Methods Nuclear magnetic resonance was used to quantify comprehensive biomarker profile in 105 146 plasma samples collected UK Biobank during 2007–2010 (age range 39–70). The were tested association with (2507 cases,...
Abstract Background The causal impact of excess adiposity on systemic metabolism is unclear. We used multivariable Mendelian randomization to compare the direct effects total (using body mass index (BMI)) and abdominal waist-to-hip-ratio (WHR)) circulating lipoproteins, lipids, metabolites with a five-fold increase in sample size over previous studies. Methods new metabolic data 109,532 UK Biobank participants. BMI WHR were measured 2006-2010, during which EDTA plasma was collected. Plasma...
Summary Pleiotropy and genetic correlation are widespread features in GWAS, but they often difficult to interpret at the molecular level. Here, we perform GWAS of 16 metabolites clustered intersection amino acid catabolism, glycolysis, ketone body metabolism a subset UK Biobank. We utilize well-documented biochemistry jointly impacting these analyze pleiotropic effects context their pathways. Among 213 lead hits, find strong enrichment for genes encoding pathway-relevant enzymes...
This paper proposes a novel method for learning highly nonlinear, multivariate functions from examples. Our takes advantage of the property that continuous can be approximated by polynomials, which in turn are representable tensors. Hence function problem is transformed into tensor reconstruction problem, an inverse decomposition. incrementally builds up unknown rank-one terms, lets us control complexity learned model and reduce chance overfitting. For models, we present efficient...
Abstract Understanding how risk factors interact to jointly influence disease can provide insights into development and improve prediction. We introduce survivalFM , a machine learning extension the widely used Cox proportional hazards model that incorporates estimation of all potential pairwise interaction effects on time-to-event outcomes. The method relies low-rank factorized approximation effects, hence overcoming computational statistical limitations fitting these terms in models...
Abstract We present comboFM , a machine learning framework for predicting the responses of drug combinations in preclinical studies, such as those based on cell lines or patient-derived cells. models context-specific interactions through higher-order tensors, and efficiently learns latent factors tensor using powerful factorization machines. The approach enables to leverage information from previous experiments performed similar drugs cells when new so far untested cells; thereby, it...
Abstract Background and Aims Early identification of individuals at high risk developing chronic kidney disease other conditions is essential for targeted prevention. Here, we assess the utility metabolic blood biomarkers in predicting onset over 250,000 from UK Biobank, beyond established factors polygenic scores. Method Circulating biomarkers, including lipids, fatty acids, amino glycolysis metabolites inflammation markers were measured by a low-cost nuclear magnetic resonance (NMR)...
Introduction: Identification of individuals at high cardiovascular risk even when using statins can guide second line therapy for primary prevention. Hypothesis: We addressed whether metabolomic biomarker profiling has utility in predicting onset major adverse events (MACE) and heart failure among already on statin therapy. Methods: Circulating lipids, fatty acids, amino glycolysis metabolites inflammation markers were measured by NMR metabolomics ~275,000 from UK Biobank. The study...
Introduction: Metabolomic biomarker profiling by Nuclear Magnetic Resonance (NMR) is a powerful technique to examine molecular signatures for heart failure risk and clarify pathological differences cardiovascular outcomes. Hypothesis: Blood lipids metabolite biomarkers may differ in how they relate prediction of vs coronary stroke Methods: We measured detailed plasma metabolites NMR metabolomics ~250,000 individuals from UK Biobank. assessed shared discordant hospitalisation relation...