A5088996741- Computational Drug Discovery Methods
- Receptor Mechanisms and Signaling
- Adenosine and Purinergic Signaling
- Protein Structure and Dynamics
- Bioinformatics and Genomic Networks
- Click Chemistry and Applications
- Metabolomics and Mass Spectrometry Studies
- Neuropeptides and Animal Physiology
- Machine Learning in Materials Science
- Molecular Biology Techniques and Applications
- Monoclonal and Polyclonal Antibodies Research
- Mass Spectrometry Techniques and Applications
- Chemical Synthesis and Analysis
- Microbial Natural Products and Biosynthesis
- Pharmacological Receptor Mechanisms and Effects
- Drug Transport and Resistance Mechanisms
- Innovative Microfluidic and Catalytic Techniques Innovation
- Neuroscience and Neuropharmacology Research
- Genetics, Bioinformatics, and Biomedical Research
- Pharmacogenetics and Drug Metabolism
- HIV/AIDS drug development and treatment
- Cardiac electrophysiology and arrhythmias
- Machine Learning in Bioinformatics
- Cannabis and Cannabinoid Research
- Chemokine receptors and signaling
Centre for Human Drug Research
2016-2025
Leiden University
2016-2025
Drug Discovery Laboratory (Norway)
2022
Wellcome Trust
2013-2017
European Bioinformatics Institute
2013-2017
KU Leuven
2013
University Hospital of Zurich
2013
The increase of publicly available bioactivity data in recent years has fueled and catalyzed research chemogenomics, mining, modeling approaches. As a direct result, over the past few multitude different methods have been reported evaluated, such as target fishing, nearest neighbor similarity-based methods, Quantitative Structure Activity Relationship (QSAR)-based protocols. However, studies are typically conducted on datasets, using validation strategies, metrics. In this study, were...
Computer-aided diagnosis of Alzheimer's disease (AD) is a rapidly developing field neuroimaging with strong potential to be used in practice. In this context, assessment models' robustness noise and imaging protocol differences together post-processing tuning strategies are key tasks addressed order move towards successful clinical applications. study, we investigated the efficacy Random Forest classifiers trained using different structural MRI measures, without neuroanatomical constraints...
Abstract Rational drug design often starts from specific scaffolds to which side chains/substituents are added or modified due the large drug-like chemical space available search for novel molecules. With rapid growth of deep learning in discovery, a variety effective approaches have been developed de novo design. In previous work we proposed method named DrugEx , can be applied polypharmacology based on multi-objective reinforcement learning. However, version is trained under fixed...
Proteochemometric modeling is founded on the principles of QSAR but able to benefit from additional information in model training due inclusion target information.
The human metabolome provides a direct physiological read-out of an individual's actual health state and includes biomarkers that may predict disease or response to treatment. discovery validation these metabolomic requires large-scale cohort studies, typically involving thousands samples. This analytical challenge drives novel technological developments enable faster, cheaper, more comprehensive analysis: for less. review summarises recent (2012–2018) towards this goal in all aspects the...
Abstract Motivation: Recent large-scale omics initiatives have catalogued the somatic alterations of cancer cell line panels along with their pharmacological response to hundreds compounds. In this study, we explored these data advance computational approaches that enable more effective and targeted use current future anticancer therapeutics. Results: We modelled 50% growth inhibition bioassay end-point (GI50) 17 142 compounds screened against 59 lines from NCI60 panel (941 831 data-points,...
Several recent studies show that inhibition of the hepatic transport proteins organic anion-transporting polypeptide 1B1 (OATP1B1) and 1B3 (OATP1B3) can result in clinically relevant drug-drug interactions (DDI). To avoid late-stage development drug failures due to OATP1B-mediated DDI, predictive vitro silico methods should be implemented at an early stage candidate evaluation process. In present study, we first developed a high-throughput transporter assay for OATP1B subfamily. A total 2000...
Proteochemometric (PCM) modelling is a computational method to model the bioactivity of multiple ligands against related protein targets simultaneously.
While a large body of work exists on comparing and benchmarking descriptors molecular structures, similar comparison protein descriptor sets is lacking. Hence, in the current total 13 different have been compared with respect to their behavior perceiving similarities between amino acids. The included study are Z-scales (3 variants), VHSE, T-scales, ST-scales, MS-WHIM, FASGAI BLOSUM, novel set termed ProtFP (4 variants). We investigate which extent show collinear as well orthogonal via...
Cyclooxygenases (COX) are present in the body two isoforms, namely: COX-1, constitutively expressed, and COX-2, induced physiopathological conditions such as cancer or chronic inflammation. The inhibition of COX with non-steroideal anti-inflammatory drugs (NSAIDs) is most widely used treatment for inflammation despite adverse effects associated to prolonged NSAIDs intake. Although selective COX-2 has been shown not palliate all (e.g. cardiotoxicity), there still niche populations which can...
Over the last 5 years deep learning has progressed tremendously in both image recognition and natural language processing. Now it is increasingly applied to other data rich fields. In drug discovery, recurrent neural networks (RNNs) have been shown be an effective method generate novel chemical structures form of SMILES. However, ligands generated by current methods so far provided relatively low diversity do not fully cover whole space occupied known ligands. Here, we propose a new (DrugEx)...
In polypharmacology drugs are required to bind multiple specific targets, for example enhance efficacy or reduce resistance formation. Although deep learning has achieved a breakthrough in de novo design drug discovery, most of its applications only focus on single target generate drug-like active molecules. However, reality molecules often interact with more than one which can have desired (polypharmacology) undesired (toxicity) effects. previous study we proposed new method named DrugEx...
Abstract With the ongoing rapid growth of publicly available ligand–protein bioactivity data, there is a trove valuable data that can be used to train plethora machine-learning algorithms. However, not all equal in terms size and quality significant portion researchers’ time needed adapt their needs. On top that, finding right for research question often challenge on its own. To meet these challenges, we have constructed Papyrus dataset. comprised around 60 million points. This dataset...
While a large body of work exists on comparing and benchmarking descriptors molecular structures, similar comparison protein descriptor sets is lacking. Hence, in the current total 13 amino acid have been benchmarked with respect to their ability establishing bioactivity models. The included study are Z-scales (3 variants), VHSE, T-scales, ST-scales, MS-WHIM, FASGAI, BLOSUM, novel set (termed ProtFP (4 variants)), addition we created three pairs combinations. Prediction performance was...
Allosteric modulators are ligands for proteins that exert their effects via a different binding site than the natural (orthosteric) ligand and hence form conceptually distinct class of target interest. Here, physicochemical structural features large set allosteric non-allosteric from ChEMBL database bioactive molecules analyzed. In general relatively smaller, more lipophilic rigid compounds, though differences exist between targets classes. Furthermore, there in distribution bind these...
Abstract The sodium taurocholate co-transporting polypeptide (NTCP, SLC10A1 ) is the main hepatic transporter of conjugated bile acids, and entry receptor for hepatitis B virus (HBV) delta (HDV). Myrcludex B, a synthetic peptide mimicking NTCP-binding domain HBV, effectively blocks HBV HDV infection. In addition, inhibits NTCP-mediated acid uptake, suggesting that also other NTCP inhibitors could potentially be novel treatment HBV/HDV This study aims to identify clinically-applied compounds...
The interpretation of high-dimensional structure–activity data sets in drug discovery to predict ligand–protein interaction landscapes is a challenging task. Here we present Drug Discovery Maps (DDM), machine learning model that maps the activity profile compounds across an entire protein family, as illustrated here for kinase family. DDM based on t-distributed stochastic neighbor embedding (t-SNE) algorithm generate visualization molecular and biological similarity. chemical target space...