M. Teunis , E. Willighagen , O. Cinar , M. Martens , M. Klaassen , M. Liem , F. Hepkema , I. Djidrovski , A. Kienhuis , C. Evelo
{"title":"S15-01 Building a virtual human for chemical risk assessment from software and data; integrating reproducible research and predictive modelling","authors":"M. Teunis , E. Willighagen , O. Cinar , M. Martens , M. Klaassen , M. Liem , F. Hepkema , I. Djidrovski , A. Kienhuis , C. Evelo","doi":"10.1016/j.toxlet.2025.07.091","DOIUrl":null,"url":null,"abstract":"<div><div>The ever-increasing scientific body of knowledge on health and disease, together with the ongoing development of innovative <em>in vitro</em> and <em>in silico</em> Non-Animal Methods (NAM), offer opportunities for animal-free safety assessment. To fully leverage the promise of NAMs and at the same time ensure safety, intersectoral and -disciplinary collaborations are necessary. This is explored in the Virtual Human Platform for Safety Assessment project (VHP4Safety). VHP4Safety is a Dutch funded research project that runs from 2021 to 2026 and brings together international partners from 32 organisations representing scientists, industry, regulators, policy makers, clinicians and Non-Animal Methods.</div><div>During the projects’ Designathons and Hackathons, data scientists, toxicologists and social scholars collaborate: (1) to build a data infrastructure to integrate existing and newly developed data and services at the Virtual Human Platform (VHP), (2) to feed the VHP with toxicological knowledge and NAM data and (3) to implement the VHP taking into account stakeholder perspectives <span><span><sup>[1]</sup></span></span>.</div><div>Here, we will demonstrate how principles of reproducible science are applied throughout the development of the VHP. We will address the most crucial technical solutions to build, test and host the VHP platform, using open science community standards. To build the VHP, an interdiciplinary development team with mixed expertise in bioinformatics, cheminformatics, molecular biology, artificial intelligence, toxicology, statistics and cloud development, work together in short sprints to deliver tangible minimal-viable products. The team has recently adopted an Agile way of working and we will share some insights on the benefits of using such an approach above the more classical ‘waterfall’ project stucture.</div><div>The VHP consists of artifacts being: software, models, documentation, standard operating procedures, workflows and data. In this presentation we will dive into how we are structuring these artifacts to form a coherent and user-friendly platform that can be used to address chemical risk-assessment questions. The VHP design and implementation is revolving around three specific case-studies: thyroid toxicity, kidney toxicity and neuro-toxicity. These case studies guide the direction of the platform development and are aimed at showcasing its capabilities. We will focus on how we are generating a sustainable platform in terms of reuse and the ability to further the develop the VHP after project end. We will showcase specific technologies that are geared towards sharing all the artifacts in the platform in a reproducible and consistent manner.</div><div><em>The VHP4Safety project is funded by the Netherlands Research Council (NWO) Netherlands Research Agenda: Research on Routes by Consortia (NWA-ORC 1292.19.272).</em></div></div>","PeriodicalId":23206,"journal":{"name":"Toxicology letters","volume":"411 ","pages":"Pages S29-S30"},"PeriodicalIF":2.9000,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Toxicology letters","FirstCategoryId":"3","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0378427425016741","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"TOXICOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
The ever-increasing scientific body of knowledge on health and disease, together with the ongoing development of innovative in vitro and in silico Non-Animal Methods (NAM), offer opportunities for animal-free safety assessment. To fully leverage the promise of NAMs and at the same time ensure safety, intersectoral and -disciplinary collaborations are necessary. This is explored in the Virtual Human Platform for Safety Assessment project (VHP4Safety). VHP4Safety is a Dutch funded research project that runs from 2021 to 2026 and brings together international partners from 32 organisations representing scientists, industry, regulators, policy makers, clinicians and Non-Animal Methods.
During the projects’ Designathons and Hackathons, data scientists, toxicologists and social scholars collaborate: (1) to build a data infrastructure to integrate existing and newly developed data and services at the Virtual Human Platform (VHP), (2) to feed the VHP with toxicological knowledge and NAM data and (3) to implement the VHP taking into account stakeholder perspectives [1].
Here, we will demonstrate how principles of reproducible science are applied throughout the development of the VHP. We will address the most crucial technical solutions to build, test and host the VHP platform, using open science community standards. To build the VHP, an interdiciplinary development team with mixed expertise in bioinformatics, cheminformatics, molecular biology, artificial intelligence, toxicology, statistics and cloud development, work together in short sprints to deliver tangible minimal-viable products. The team has recently adopted an Agile way of working and we will share some insights on the benefits of using such an approach above the more classical ‘waterfall’ project stucture.
The VHP consists of artifacts being: software, models, documentation, standard operating procedures, workflows and data. In this presentation we will dive into how we are structuring these artifacts to form a coherent and user-friendly platform that can be used to address chemical risk-assessment questions. The VHP design and implementation is revolving around three specific case-studies: thyroid toxicity, kidney toxicity and neuro-toxicity. These case studies guide the direction of the platform development and are aimed at showcasing its capabilities. We will focus on how we are generating a sustainable platform in terms of reuse and the ability to further the develop the VHP after project end. We will showcase specific technologies that are geared towards sharing all the artifacts in the platform in a reproducible and consistent manner.
The VHP4Safety project is funded by the Netherlands Research Council (NWO) Netherlands Research Agenda: Research on Routes by Consortia (NWA-ORC 1292.19.272).