Georgia Tsiliki

We present toxFlow, a web-application developed for enrichment analysis of omics data coupled with read-across toxicity prediction. A sequential analysis workflow is suggested where users can filter omics data using enrichment scores and incorporate their findings into a correlation-based read-across technique for predicting the toxicity of a substance based on its analogs. Either embedded or in-house gene signature libraries can be used for enrichment analysis. The suggested approach can be used for toxicity prediction of diverse chemical entities, however this article focuses on the multi-perspective characterization of nanoparticles and selects their neighbors based on both physicochemical and biological similarity criteria. In addition, visualization options are offered to interactively explore correlation patterns in the data, whereas results can be exported for further analysis. toxFlow is accessible at http://147.102.86.129:3838/toxflow.

The interest toward omics data is growing in the field of toxicology owing to the diverse knowledge they generate, which can improve prediction and dosage profiling for more accurate safety assessment. An integration methodology is presented where high‐throughput omics data are enriched with biological‐pathway information to produce a novel set of biological (BIO) descriptors by decomposing omics data to meaningful clusters in terms of both their mechanistic interpretation and correlation affinity. A generalized simulated annealing algorithm is employed to estimate the optimal partition of the enriched data and accordingly produce novel descriptors based on gene content similarity. BIO descriptors are characterized by the pathway information fused to the data; thereby, they refer to groups of genes with similar biological implications rather than specific genes, which could vary across studies. The methodology is applied to an extensive proteomics data set and demonstrates that BIO descriptors are beneficial for modeling prediction, outperforming the prediction accuracy of the original omics data, and offering a readily available biological interpretation of the findings.

Engineered nanomaterials (ENMs) are increasingly infiltrating our lives as a result of their applications across multiple fields. However, ENM formulations may result in the modulation of pathways and mechanisms of toxic action that endanger human health and the environment. Alternative testing methods such as in silico approaches are becoming increasingly popular for assessing the safety of ENMs, as they are cost- and time-effective. Additionally, computational approaches support the industrial safer-by-design challenge and the REACH legislation objective of reducing animal testing. Because of the novelty of the field, there is also an evident need for harmonization in terms of databases, ontology, and modeling infrastructures. To this end, we present Jaqpot Quattro, a comprehensive open-source web application for ENM modeling with emphasis on predicting adverse effects of ENMs. We describe the system architecture and outline the functionalities, which include nanoQSAR modeling, validation services, read-across predictions, optimal experimental design, and interlaboratory testing.

Background: The NanoSafety Cluster, a cluster of projects funded by the European Commision, identified the need for a computational infrastructure for toxicological data management of engineered nanomaterials (ENMs). Ontologies, open standards, and interoperable designs were envisioned to empower a harmonized approach to European research in nanotechnology. This setting provides a number of opportunities and challenges in the representation of nanomaterials data and the integration of ENM information originating from diverse systems. Within this cluster, eNanoMapper works towards supporting the collaborative safety assessment for ENMs by creating a modular and extensible infrastructure for data sharing, data analysis, and building computational toxicology models for ENMs.

The term ‘molecular cartography’ encompasses a family of computational methods for two-dimensional transformation of protein structures and analysis of their physicochemical properties. The underlying algorithms comprise multiple manual steps, whereas the few existing implementations typically restrict the user to a very limited set of molecular descriptors. We present Structuprint, a free standalone software that fully automates the rendering of protein surface maps, given - at the very least - a directory with a PDB file and an amino acid property. The tool comes with a default database of 328 descriptors, which can be extended or substituted by user-provided ones. The core algorithm comprises the generation of a mould of the protein surface, which is subsequently converted to a sphere and mapped to two dimensions, using the Miller cylindrical projection. Structuprint is partly optimized for multicore computers, making the rendering of animations of entire molecular dynamics simulations feasible. Structuprint is an efficient application, implementing a molecular cartography algorithm for protein surfaces. According to the results of a benchmark, its memory requirements and execution time are reasonable, allowing it to run even on low-end personal computers. We believe that it will be of use - primarily but not exclusively - to structural biologists and computational biochemists.

Cheminformatics and bioinformatics are extensively using predictive modelling and exhibit a need for standardization of methodologies such as data splitting, cross-validation methods, best model criteria and Y-randomization. RRegrs is a new R package, available at https://www.github.com/enanomapper/RRegrs (0.05 release), which suggests an integrated framework to assist model selection and speed up the process of predictive model development. The tool proposes a fully validated scheme by employing repeated 10-fold and leave-one-out cross-validation for ten linear and non-linear regression methods. Standardized reports are produced to compare the output of modelling algorithms and assess cross-validation results for selected models. Here, we demonstrate RRegrs capabilities in terms of performance using five well-established data sets.

Predictive regression models can be created with many different modelling approaches. Choices need to be made for data set splitting, cross-validation methods, specific regression parameters and best model criteria, as they all affect the accuracy and efficiency of the produced predictive models, and therefore, raising model reproducibility and comparison issues. Cheminformatics and bioinformatics are extensively using predictive modelling and exhibit a need for standardization of these methodologies in order to assist model selection and speed up the process of predictive model development. A tool accessible to all users, irrespectively of their statistical knowledge, would be valuable if it tests several simple and complex regression models and validation schemes, produce unified reports, and offer the option to be integrated into more extensive studies. Additionally, such methodology should be implemented as a free programming package, in order to be continuously adapted and redistributed by others. We propose an integrated framework for creating multiple regression models, called RRegrs. The tool offers the option of ten simple and complex regression methods combined with repeated 10-fold and leave-one-out cross-validation. Methods include Multiple Linear regression, Generalized Linear Model with Stepwise Feature Selection, Partial Least Squares regression, Lasso regression, and Support Vector Machines Recursive Feature Elimination. The new framework is an automated fully validated procedure which produces standardized reports to quickly oversee the impact of choices in modelling algorithms and assess the model and cross-validation results. The methodology was implemented as an open source R package, available at https://www.github.com/enanomapper/RRegrs , by reusing and extending on the caret package. The universality of the new methodology is demonstrated using five standard data sets from different scientific fields. Its efficiency in cheminformatics and QSAR modelling is shown with three use cases: proteomics data for surface-modified gold nanoparticles, nano-metal oxides descriptor data, and molecular descriptors for acute aquatic toxicity data. The results show that for all data sets RRegrs reports models with equal or better performance for both training and test sets than those reported in the original publications. Its good performance as well as its adaptability in terms of parameter optimization could make RRegrs a popular framework to assist the initial exploration of predictive models, and with that, the design of more comprehensive in silico screening applications.

Background:The NanoSafety Cluster, a cluster of projects funded by the European Commision, identified the need for a computational infrastructure for toxicological data management of engineered nanomaterials (ENMs). Ontologies, open standards, and interoperable designs were envisioned to empower a harmonized approach to European research in nanotechnology. This setting provides a number of opportunities and challenges in the representation of nanomaterials data and the integration of ENM information originating from diverse systems. Within this cluster, eNanoMapper works towards supporting the collaborative safety assessment for ENMs by creating a modular and extensible infrastructure for data sharing, data analysis, and building computational toxicology models for ENMs.Results:The eNanoMapper database solution builds on the previous experience of the consortium partners in supporting diverse data through flexible data storage, open source components and web services. We have recently described the design of the eNanoMapper prototype database along with a summary of challenges in the representation of ENM data and an extensive review of existing nano-related data models, databases, and nanomaterials-related entries in chemical and toxicogenomic databases. This paper continues with a focus on the database functionality exposed through its application programming interface (API), and its use in visualisation and modelling. Considering the preferred community practice of using spreadsheet templates, we developed a configurable spreadsheet parser facilitating user friendly data preparation and data upload. We further present a web application able to retrieve the experimental data via the API and analyze it with multiple data preprocessing and machine learning algorithms.Conclusion:We demonstrate how the eNanoMapper database is used to import and publish online ENM and assay data from several data sources, how the “representational state transfer” (REST) API enables building user friendly interfaces and graphical summaries of the data, and how these resources facilitate the modelling of reproducible quantitative structure–activity relationships for nanomaterials (NanoQSAR).

The EU-funded eNanoMapper project proposes a computational infrastructure for toxicological data management of engineered nanomaterials (ENMs) based on open standards, ontologies and an interoperable design to enable a more effective, integrated approach to European research in nanotechnology. eNanoMapper's goal is to support the collaborative safety assessment for ENMs by creating a modular, extensible infrastructure for transparent data sharing, data analysis, and the creation of computational toxicology models for ENMs. The eNanoMapper database solution builds on previous experience of the consortium partners in supporting diverse data through flexible data storage, semantic web technologies, open source components and web services. A number of opportunities and challenges exist in nanomaterials representation and integration of ENM information, originating from diverse systems. A short summary, highlighting the pros and cons of the existing integration approaches and data models is presented. We demonstrate the approach of adopting an ontology-supported data model, describing the materials and measurements. The data sources supported include diverse formats (ISA-Tab, OECD Harmonized Templates, custom spreadsheet templates, various databases provided by consortia members). Besides retaining the data provenance, the focus on measurements provides insights into how to reuse the chemical structure database tools for nanomaterials characterization and safety.