A5046035604- Computational Drug Discovery Methods
- Machine Learning in Materials Science
- Analytical Chemistry and Chromatography
- Pharmacogenetics and Drug Metabolism
- Machine Learning and Data Classification
- Animal testing and alternatives
- Chemistry and Chemical Engineering
- Carcinogens and Genotoxicity Assessment
- Manufacturing Process and Optimization
- Statistical Methods in Clinical Trials
- Machine Learning and Algorithms
- Pharmacovigilance and Adverse Drug Reactions
- Microbial Metabolism and Applications
- Effects and risks of endocrine disrupting chemicals
- Maritime Ports and Logistics
- Statistical Methods and Inference
- Metabolomics and Mass Spectrometry Studies
- Machine Learning in Bioinformatics
- Time Series Analysis and Forecasting
- Occupational exposure and asthma
- Pesticide and Herbicide Environmental Studies
- Dermatology and Skin Diseases
- Drug-Induced Hepatotoxicity and Protection
- Domain Adaptation and Few-Shot Learning
- Melanoma and MAPK Pathways
Uppsala University
2020-2024
Science for Life Laboratory
2021-2024
Royal Holloway University of London
2023
AstraZeneca (Sweden)
2012-2021
Stena (Sweden)
2021
AstraZeneca (Brazil)
2018
The present publication surveys several applications of in silico (i.e., computational) toxicology approaches across different industries and institutions. It highlights the need to develop standardized protocols when conducting toxicity-related predictions. This contribution articulates information needed for support predictions major toxicological endpoints concern (e.g., genetic toxicity, carcinogenicity, acute reproductive developmental toxicity) regulatory bodies. Such novel (IST)...
In silico toxicology (IST) approaches to rapidly assess chemical hazard, and usage of such methods is increasing in all applications but especially for regulatory submissions, as assessing chemicals under REACH well the ICH M7 guideline drug impurities. There are a number obstacles performing an IST assessment, including uncertainty how assessment associated expert review should be performed or what fit purpose, lack confidence that results will accepted by colleagues, collaborators...
Acid–base properties of molecules in nonaqueous solvents are critical importance for almost all areas chemistry. Despite this very high relevance, our knowledge is still mostly limited to the pKa rather few compounds most common solvents, and a simple yet truly general computational procedure predict pKa's any compound solvent missing. In contribution, we describe such procedure. Our method requires only experimental reference water standard quantum-chemical calculations. This tested through...
Conformal prediction has been proposed as a more rigorous way to define confidence compared other application domain concepts that have earlier used for QSAR modeling. One main advantage of such method is it provides region potentially with multiple predicted labels, which contrasts the single valued (regression) or label (classification) output predictions by standard modeling algorithms. Standard conformal might not be suitable imbalanced data sets. Therefore, Mondrian cross-conformal...
Abstract Conformal prediction has seen many applications in pharmaceutical science, being able to calibrate outputs of machine learning models and producing valid intervals. We here present the open source software CPSign that is a complete implementation conformal for cheminformatics modeling. implements inductive transductive classification regression, probabilistic with Venn-ABERS methodology. The main chemical representation signatures but other types descriptors are also supported....
The assessment of skin sensitization has evolved over the past few years to include in vitro assessments key events along adverse outcome pathway and opportunistically capitalize on strengths silico methods support a weight evidence without conducting test animals. While vary greatly their purpose format; there is need standardize underlying principles which such models are developed make transparent implications for uncertainty overall assessment. In this contribution, relationship between...
ABSTRACT Purpose Pharmacovigilance methods have advanced greatly during the last decades, making post‐market drug assessment an essential evaluation component. These mainly rely on use of spontaneous reporting systems and health information databases to collect expertise from huge amounts real‐world reports. The EU‐ADR Web Platform was built further facilitate accessing, monitoring exploring these data, enabling in‐depth analysis adverse reactions risks. Methods exploits wealth data...
Many drugs designed to inhibit kinases have their clinical utility limited by cardiotoxicity-related label warnings or prescribing restrictions. While this liability is widely recognized, designing safer kinase inhibitors (KI) requires knowledge of the causative kinase(s). Efforts unravel encountered pharmacology with nearly prohibitive complexity. At therapeutically relevant concentrations, KIs show promiscuity distributed across kinome. Here, overcome complexity, 65 known kinome-scale...
Ligand-based models can be used in drug discovery to obtain an early indication of potential off-target interactions that could linked adverse effects. Another application is combine such into a panel, allowing compare and search for compounds with similar profiles. Most contemporary methods implementations however lack valid measures confidence their predictions, only provide point predictions. We here describe methodology uses Conformal Prediction predicting interactions, trained on data...
Structural alerts have been one of the backbones computational toxicology and applications in many areas including cosmetic, environmental, pharmaceutical toxicology. The development structural has always involved a manual analysis existing data related to relevant end point followed by determination substructures that appear be specific outcome. are then analyzed for their utility posterior validation studies, which at times stretched over years or even decades. With higher throughput...
<p>Each year the pharmaceutical industry makes thousands of compounds, many which do not meet desired efficacy or pharmacokinetic properties, describing absorption, distribution, metabolism and excretion (ADME) behavior. Parameters such as lipophilicity, solubility metabolic stability can be measured in high throughput vitro assays. However, a compound needs to synthesized order tested. In silico models for these endpoints exist, although with varying quality. Such used before...
The occurrence of mutagenicity in primary aromatic amines has been investigated using conformal prediction. results the investigation show that it is possible to develop mathematically proven valid models prediction and existence uncertain classes prediction, such as both (both assigned a compound) empty (no class compound), provides user with additional information on how use, further develop, possibly improve future models. study also indicates use different sets fingerprints models, for...
Pharmacovigilance plays a key role in the healthcare domain through assessment, monitoring and discovery of interactions amongst drugs their effects human organism. However, technological advances this field have been slowing down over last decade due to miscellaneous legal, ethical methodological constraints. Pharmaceutical companies started realize that collaborative integrative approaches boost current drug research development processes. Hence, new strategies are required connect...
Valid and predictive models for classifying Ames mutagenicity have been developed using conformal prediction. The are Random Forest signature molecular descriptors. investigation indicates, on excluding not-strongly mutagenic compounds (class B), that the validity is increased predictions based both public Division of Genetics Mutagenesis, National Institute Health Sciences Japan (DGM/NIHS) data while less so when only latter source. former result in valid majority, non-mutagenic, class...
Loading and unloading rolling cargo in roll-on/roll-off are important very recurrent operations maritime logistics. In this paper, we apply state-of-the-art deep reinforcement learning algorithms to automate these a complex real environment. The objective is teach an autonomous tug master manage perform loading while avoiding collisions with static dynamic obstacles along the way. artificial intelligence agent, representing master, trained evaluated challenging environment based on Unity3D...
We argue for supplementing the process of training a prediction algorithm by setting up scheme detecting moment when distribution data changes and needs to be retrained. Our proposed schemes are based on exchangeability martingales, i.e., processes that martingales under any exchangeable data. method, conformal prediction, is general can applied top modern algorithm. Its validity guaranteed, in this paper we make first steps exploring its efficiency.
The drug-induced accumulation of phospholipids in lysosomes various tissues is predominantly observed regular repeat dose studies, often after prolonged exposure, and further investigated mechanistic studies prior to candidate nomination. finding can cause delays the discovery process inflicting high costs affected projects. This article presents an vitro imaging-based method for early detection phospholipidosis liability a hybrid approach risk mitigation phospolipidosis utilizing readout...