Data Services

About

[1]

Data science is the science of data management based on computer science and mathematical, statistical, machine learning and artificial intelligence methods supporting the use of in silico approaches to guide decision making and strategic planning. We practice such approaches in the development of applications supporting product design, predictive toxicology and risk assessment.

Data collection

During this step, data is collected. Data might be collected out of databases, documents, real time streaming or from ongoing research processes and exist in a lot of different formats including image files, raw data, processed data and results.

Data storage and processing

This step includes the processing of data by data management teams which set the standards for data storage. Data is stored in repositories where it can be organized, searched and accessed.

Data analysis

During data analysis, data is retrieved from repositories and integrated with analysis procedures, machine learning, artificial intelligence and other algorithms so that it can be used for predictive modeling and risk assessment.

Reporting

In the final step an analysis or assessment report is generated with explanations and visualization of the analysis supported by referenced protocols and data sources of high integrity.

Reference: > IBM

[1] Image from Advanced tools for the safety assessment of nanomaterials, by fadeel B., Farcal L., Hardy B., Vàzquez-Campos S., Hristozov D., Marcomini A., Lynch I., Valsami-Jones E., Alenius H. and Savolainen K. in Nature Nanotechnology 13, 537-543, DOI: 10.1038/s41565-018-0185-0, 2018.

Our data services

[1]

With our packages we contribute to the early stage risk assessment of chemicals for different concerns including systemic and organ toxicity, neurotoxicity, skin irritation and sensitization, endocrine disruption and environmental endpoints. The computational approach includes:

• Integrating data on physical and chemical properties such as solubility, reactivity, volatility, biodegradability etc.

• Integrating (eco)toxicological data from standardized or non-standardized tests.

• Integrated in silico approaches where we apply machine learning algorithms to evaluate compounds based on combinations of chemical and biological features

• Read across for a compound grouping (based on best practices as developed under EU-ToxRisk and interactions with OECD/ECHA).

Therefore, it is important to consider several main elements throughout the data processing workflow and to have a clear management of data and knowledge life cycle. This includes the definition of how the data should be structured; for example, data inputs can be defined over the input panels of a software application where it is made clear what data is necessary to be given as input to be able to further process it. There are several methods to import data and it has to be evaluated in each case what method should be used to be able to process it as efficiently as possible. Data can directly be processed by accessing it from a database or by importing it from a template file into the corresponding data analysis tools.

Application of our data services using our integrated solutions include:

• SaferNeuron: BBB modelling, target and off-target interactions.
• SaferSkin: Bayesian Network-based IATAs, determination of potencies.
• SaferEndocrine: QSARs, Species-distributions, developing and using EATS models.
• SaferLiver: Constructing and implementing tiered strategies and models for cholestasis, steatosis, DILI.

[1] Reference: Image based and adapted from Data Flair (> Link), March 2023.

Partner Contributions

Beside our contribution we offer further approaches together with our partners. Solutions include:

Kreatis

Keratis offers QSAR data modeling which can be accessed by iSafeRat. iSafeRat is a user-friendly platform where you can run MechoA scheme models, predict mechanisms of toxic action of substances in various species and design safe chemicals (ecoconception).

> Read more

IRFMN

The Instituto di Ricerche Pharmacologiche Mario Negri (IRFMN) provides expertise in computational predictive toxicology based on QSAR implementation. IRFMN puts great emphasis on ensuring that their models generate transparent, understandable, reproducible and verifiable results.

> Read more

Euclia

Euclia provides you with the Jaqpot, a cloud-native platform that offers an easy and efficient way of developing, using, sharing, and integrating predictive models at scale, in a fully documented way.

> Read more

Maieutix

empower small and middle-size biotech companies and drug makers with innovative technologies and to support the development of innovative project within the pharma environment.

> Read more

Multicase

Multicase creates commercially viable products through the development of innovative (Q)SAR methodologies and software solutions.

> Read more

Phenaris

Phenaris uses cutting-edge technologies and provides data, models, and decision support in all aspects of in silico toxicology. They are currently working on multiple solutions to provide data integration, model development, and decision support.

> Read more

SafeDose Ltd.

SafeDose Ltd. helps its clients to see their way through the maze. How, you ask? By understanding their client’s current risk assessment processes and helping improve them.

> Read more

SBX Corporation

SBX Corporation provides solutions to complex socio-industrial problems in biology and beyond, using Artificial Intelligence/Machine Learning, Internet of Things/sensor network, and model-based simulation techniques.

> Read more

SciLicium

SciLicium is specialized in multi-omics data analysis and the development of workflows and web app solutions in bioinformatics. Over the last few years, SciLicium has also developed strong expertise in toxicogenomics (read-across, adverse outcome pathways, mechanisms of action) including predictive toxicogenomics based on statistical modeling (Machine and Deep Learning).

> Read more

The Vertex laboratories

The Vertex team’s expertise in the areas of Machine Learning, Deep Learning, Automation and Data Science allows them to provide custom Artificial Intelligence solutions for clients, as opposed to off-the-shelf solutions.

> Read more

What is EdelweissData™?

EdelweissData™ is a platform that solves your problems with sharing, combining and integrating data. EdelweissData™ prepares, curates, harmonises, standardises and brings together data from different formats, different sources, and different labs.

Our solution is:

• cloud-based,
• platform independent, and
• highly customisable with respect to data formats and workflows.

EdelweissData™ provides applications to support:

• importation, harmonisation, annotation and storage of primary data and metadata,
• storage in agreement with FAIR data sharing principles,
• user-friendly, non-expert level searching and browsing, as well as analysis and visualisation, both interactively and in scripting mode,
• options for customisation to address specific project requirements including optimal interplay with other existing and newly developed infrastructure.

EdelweissData™ was developed and is currently used and marketed within the context of predictive toxicology and safety assessment use cases, but can be deployed to solve related issues in adjacent Life Science areas.

Why EdelweissData™?

EdelweissData™ helps you discover valuable data connections from different sources, integrate it with your own data and share the enhanced resource. EdelweissData™ helps you find and annotate the best information that resources in a targeted field have to offer, putting it to best use for your application and user needs.

What EdelweissData™ can do for you

• Integrate
• Harmonise
• Manage
• Automatically upload data for improved accuracy
• Suggest new directions
• Make your science reproducible

How EdelweissData™ grew from Edelweiss Connect's experience

Edelweiss Connect brings together an unusual combination of experts in data science, life sciences and chemistry. Edelweiss Connect earned its reputation by providing powerful data management solutions to academic and industry partners in such major collaborative projects as:

INTERVALS™: Edelweiss Connect (formerly Douglas Connect) developed the online platform and associated data warehouse of the INTERVALS project initiated by Philip Morris International R&D. INTERVALS™ aims to develop the science for a smoke-free world by sharing protocols, tools and data from assessment studies, thus enabling independent third-party collaboration and data analysis.

EU-ToxRisk: Edelweiss Connect customised EdelweissData™ to respond to the challenges of the large, complex EU-ToxRisk project, an integrated European ‘Flagship’ programme to drive mechanism-based toxicity testing and risk assessment for the 21st century.

> Link

ACEnano: Edelweiss Connect customised EdelweissData™ to provide a knowledge warehouse for the complex protocols, metadata and data required to characterise nanotechnology. ACEnano aims to introduce confidence, adaptability and clarity into nanomaterial risk assessment.

> Link

OpenTox: Edelweiss Connect helped create and lead OpenTox, the leading global open platform for predictive toxicology supported by open standards.

> Link

NanoCommons: Edelweiss Connect developed a public version of the ACEnano warehouse for nanomaterial characterization as part of the all-encompassing nanosafety data universe created in the H2020 Infrastructures project NanoCommons. By creating a community framework and infrastructure for reproducible science, and in particular, in silico workflows for nanomaterials and beyond, NanoCommons seeks to address the need for nanosafety knowledge management.

> Link

Features

Based on your requirements and the personalised set-up of EdelweissData™ that is deployed, you could enjoy the following features and benefits.

Life Sciences data management

• A comprehen­sive tabular data and metadata environment
• Supports annotation, organisation and storage of primary data and metadata
• Provides domain data types (e.g. understand chemical’s SMILES)
• Facilitates the analysis, visualisation and sharing of data
• Provides interactive exploration of the data via web-based tools
• Implements the FAIR data principles of Findability, Accessibility, Interoperability and Reusability.
• Allows the upload of data directly onto a secure, cloud-based platform
• Provides harmonised and interoperable access to different knowledge sources including publicly available databases
• Provides a rich application programming interface (API)
• Helps creating a culture of data sharing by making sharing easy
• Replace manual error-prone, time consuming and costly processes with lean data solutions and processing workflows.

Deployment and user experience

• Simple and functional
• Cloud-based or deployed in-house
• Hides the complexity of managing the database
• Accessible from PCs, tablets and smartphones
• Suitable for diverse local and virtual teams of expert and non-expert users alike
• Designed to build on top of and integrate with the existing data storage systems and data workflows, rather than replacing them (but could work standalone if needed)
• Capable of integrating with data-producing hardware such as analytical instruments, pre-existing locally stored data, and data sources available on the internet
• Provides flexible data access policies across tools, APIs, users and groups

EdelweissData™ Tool

EdelweissData™ is an exciting software solution that offers scientists seamless sharing of data and metadata.

> EdelweissData™ Tool

ToxCalc!

SaferWorldbyDesign provider SafeDose Ltd. is a Canadian human health risk assessment consultancy and software development company. The team at SafeDose recognizes that chemical safety assessments can be challenging, involving a wide range of data, regulations, and assessment methods. Despite this complexity, they are committed to guiding clients through the process and ensuring the development of high-quality assessments.

About ToxCalc!

ToxCalc! is a website purpose-built for toxicologists and other scientific professionals to help us do routine calculations. The goal of ToxCalc! is to help those of us who assess chemical hazards and risk work faster, with fewer errors, using the best available evidence. ToxCalc! is also designed to promote the principles of systematic review. It wants you to record and reference the calculations and default values you used and append them to your chemical safety reports so that everyone can see where the numbers came from. Finally, ToxCalc! is also designed to help you understand the math and biology behind the calculations through the use of tutorials and help features.

What ToxCalc! currently does

Currently, ToxCalc! implements five multidirectional calculators (converts between doses and concentrations, animal and human doses, corrects for bioavailability, and calculates health-based exposure limits).

> Try the calculators now

How to contribute?

We would love to have you contribute to this project!

If you are a toxicologist or scientific professional working in chemical safety assessment, we’d welcome your suggestions for calculations. We will prioritize calculations that are most useful to a broad audience, published in peer-reviewed journals or regulatory guidelines, and validated. Please cite the URL, or preferably, the published reference when submitting your request. If your calculation is accepted, you will be recognized as a ToxCalc! contributor. Another way to contribute is to act as a peer reviewer. Peer reviewers check the accuracy and relevance of the calculators and tools within ToxCalc!. If you’d like to be a peer reviewer, let us know here.

Are you a software developer, expert in math, visual design, usability, or any other discipline that ToxCalc! could benefit from? Get in touch here. You may earn coursework or certification credits by working on this project. 

If you are interested in our developing and deploying such resources to support your work, please contact us to discuss your requirements and obtain a proposal.