Showing papers by "Nina Jeliazkova published in 2020"
Research Triangle Park1, National Institutes of Health2, Technische Universität München3, University of Bari4, University of North Carolina at Chapel Hill5, Universidade Federal de Goiás6, Umeå University7, Lanzhou University8, University of Milano-Bicocca9, Mario Negri Institute for Pharmacological Research10, University of Insubria11, Karolinska Institutet12, Dalian University of Technology13, North Carolina State University14, University of Tartu15, Food and Drug Administration16, University of Strasbourg17, United States Environmental Protection Agency18, Technical University of Denmark19, University of Paris20, Institute of Business & Medical Careers21, U.S. Agency for Toxic Substances and Disease Registry22, University of Chicago23, East China University of Science and Technology24
TL;DR: The Collaborative Modeling Project for Androgen Receptor Activity (CoMPARA) efforts are described, which follows the steps of the Collaborative Estrogen Recept Activity Prediction Project (CERAPP).
Abstract: BACKGROUND: Endocrine disrupting chemicals (EDCs) are xenobiotics that mimic the interaction of natural hormones and alter synthesis, transport, or metabolic pathways. The prospect of EDCs causing ...
107 citations
TL;DR: The GRACIOUS Framework provides an initial set of hypotheses for the grouping of NFs which take into account the identity and use of the NFs, as well as the purpose of grouping, to facilitate the application of grouping of nanomaterials or nanoforms in a regulatory context and to support innovation.
71 citations
TL;DR: This is the first large-scale study evaluating the potential of machine learning and especially deep learning directly at the level of industry-scale settings and moreover investigating the transferability of publicly learned target prediction models towards industrial bioactivity prediction pipelines.
Abstract: Artificial intelligence (AI) is undergoing a revolution thanks to the breakthroughs of machine learning algorithms in computer vision, speech recognition, natural language processing and generative modelling. Recent works on publicly available pharmaceutical data showed that AI methods are highly promising for Drug Target prediction. However, the quality of public data might be different than that of industry data due to different labs reporting measurements, different measurement techniques, fewer samples and less diverse and specialized assays. As part of a European funded project (ExCAPE), that brought together expertise from pharmaceutical industry, machine learning, and high-performance computing, we investigated how well machine learning models obtained from public data can be transferred to internal pharmaceutical industry data. Our results show that machine learning models trained on public data can indeed maintain their predictive power to a large degree when applied to industry data. Moreover, we observed that deep learning derived machine learning models outperformed comparable models, which were trained by other machine learning algorithms, when applied to internal pharmaceutical company datasets. To our knowledge, this is the first large-scale study evaluating the potential of machine learning and especially deep learning directly at the level of industry-scale settings and moreover investigating the transferability of publicly learned target prediction models towards industrial bioactivity prediction pipelines.
34 citations
Norwegian Institute for Air Research1, DECHEMA2, University of Chester3, University of Bergen4, Bergen University College5, Swansea University6, ANSES7, Health Canada8, Katholieke Universiteit Leuven9, Engineer Research and Development Center10, University of Birmingham11, Spanish National Research Council12, Federal Institute for Occupational Safety and Health13, Procter & Gamble14, University of Gdańsk15, Amity University16, Wilmington University17, Duke University18, Maastricht University19
TL;DR: This paper provides an overview of the state-of-the-art regarding risk governance of NMs and lays out the theoretical basis for the development and implementation of an effective, trustworthy and transparent risk governance framework for NMs.
Abstract: Nanotechnologies have reached maturity and market penetration that require nano-specific changes in legislation and harmonization among legislation domains, such as the amendments to REACH for nanomaterials (NMs) which came into force in 2020. Thus, an assessment of the components and regulatory boundaries of NMs risk governance is timely, alongside related methods and tools, as part of the global efforts to optimise nanosafety and integrate it into product design processes, via Safe(r)-by-Design (SbD) concepts. This paper provides an overview of the state-of-the-art regarding risk governance of NMs and lays out the theoretical basis for the development and implementation of an effective, trustworthy and transparent risk governance framework for NMs. The proposed framework enables continuous integration of the evolving state of the science, leverages best practice from contiguous disciplines and facilitates responsive re-thinking of nanosafety governance to meet future needs. To achieve and operationalise such framework, a science-based Risk Governance Council (RGC) for NMs is being developed. The framework will provide a toolkit for independent NMs' risk governance and integrates needs and views of stakeholders. An extension of this framework to relevant advanced materials and emerging technologies is also envisaged, in view of future foundations of risk research in Europe and globally.
20 citations
TL;DR: A workflow to facilitate the conversion of spreadsheets into a FAIR (Findable, Accessible, Interoperable, and Reusable) database, with the pivotal aid of the NMDataParser tool, developed to streamline the mapping of the original file layout into the eNanoMapper semantic data model.
Abstract: The field of nanoinformatics is rapidly developing and provides data driven solutions in the area of nanomaterials (NM) safety. Safe by Design approaches are encouraged and promoted through regulatory initiatives and multiple scientific projects. Experimental data is at the core of nanoinformatics processing workflows for risk assessment. The nanosafety data is predominantly recorded in Excel spreadsheet files. Although the spreadsheets are quite convenient for the experimentalists, they also pose great challenges for the consequent processing into databases due to variability of the templates used, specific details provided by each laboratory and the need for proper metadata documentation and formatting. In this paper, we present a workflow to facilitate the conversion of spreadsheets into a FAIR (Findable, Accessible, Interoperable, and Reusable) database, with the pivotal aid of the NMDataParser tool, developed to streamline the mapping of the original file layout into the eNanoMapper semantic data model. The NMDataParser is an open source Java library and application, making use of a JSON configuration to define the mapping. We describe the JSON configuration syntax and the approaches applied for parsing different spreadsheet layouts used by the nanosafety community. Examples of using the NMDataParser tool in nanoinformatics workflows are given. Challenging cases are discussed and appropriate solutions are proposed.
13 citations
08 Apr 2020
TL;DR: A continuously updated dataset, METER.AC is described, consisting of raw measurements of atmospheric pressure, temperature, relative humidity, particulate matter, and background radiation in about 100 locations in Bulgaria, as well as some derived values such as sea-level atmosphericpressure, dew/frost point, and hourly trends.
Abstract: Detailed atmospheric monitoring data are notoriously difficult to obtain for some geographic regions, while they are of paramount importance in scientific research, forecasting, emergency response, policy making, etc. We describe a continuously updated dataset, METER.AC, consisting of raw measurements of atmospheric pressure, temperature, relative humidity, particulate matter, and background radiation in about 100 locations in Bulgaria, as well as some derived values such as sea-level atmospheric pressure, dew/frost point, and hourly trends. The measurements are performed by low-power maintenance-free nodes with common hardware and software, which are specifically designed and optimized for this purpose. The time resolution of the measurements is 5 min. The short-term aim is to deploy at least one node per 100 km2, while uniformly covering altitudes between 0 and 3000 m asl with a special emphasis on remote mountainous areas. A full history of all raw measurements (non-aggregated in time and space) is publicly available, starting from September 2018. We describe the basic technical characteristics of our in-house developed equipment, data organization, and communication protocols as well as present some use case examples. The METER.AC network relies on the paradigm of the Internet of Things (IoT), by collecting data from various gauges. A guiding principle in this work is the provision of findable, accessible, interoperable, and reusable (FAIR) data. The dataset is in the public domain, and it provides resources and tools enabling citizen science development in the context of sustainable development.
6 citations