New, Free AI Tool Helps Assess Manuscripts for Animal Research Transparency

A new and freely available tool will streamline manuscript checks to ensure that animal research is transparently reported in line with ARRIVE Essential 10.

The freely available compliance checker will use natural language processing to automatically assess scientific manuscripts for the information in the ARRIVE Essential 10, a checklist of the most important details to include in any publication describing animal research.

A major factor influencing the reliability and reproducibility of animal experiments is how transparently they are reported. The ARRIVE guidelines help researchers improve the reporting of animal studies by clearly laying out the information that should be included in a manuscript. Reporting animal experiments in line with the ARRIVE guidelines is a requirement of research funding organizations worldwide, including all major UK animal research funders, and the National Institutes of Health in the US.

Assessing a manuscript’s compliance with ARRIVE is a valuable way for researchers and journals to identify when key information is missing from it. Presently, ARRIVE compliance assessment can only be performed manually, using the ARRIVE compliance questionnaire. Automating this process will facilitate the inclusion of ARRIVE assessments in journals’ manuscript workflows, as well as the auditing of reporting practices by funding organizations or universities.

The compliance checker will be developed by the ScreenIT consortium, a group of developers led by SciCrunch, a California-based company focused on initiatives to improve research reproducibility. SciCrunch develop and operate platforms such as the Resource Identification Initiative, the Antibody Registry, and SciScore, a popular scientific manuscript review tool that uses the same technology that will be implemented in the ARRIVE compliance checker.

Nathalie Percie du Sert, PhD, head of experimental design and reporting at the NC3Rs, said:

“We are delighted to work with SciCrunch and colleagues to develop such an important tool for the research community. The automated ARRIVE checker will provide researchers, journals, and reviewers with a freely accessible tool to check that manuscripts are not missing critical information and it will allow us and other stakeholders to monitor improvements in the reporting of animal research.”

Anita Bandrowski, PhD, CEO of SciCrunch, said:

“The ARRIVE guidelines are a very important tool for improving the reporting of animal research, we are very pleased to be developing the compliance checker, which will help make their use easier for the community. At the same time, we’re glad to be able to take this opportunity to add ARRIVE compliance checks to SciScore, making ARRIVE assessment functionality available to the large number of journals who already use SciScore."

The compliance checker will be available via the ARRIVE guidelines website in 2025. Before that, ARRIVE compliance checking functionality will be added to SciScore in 2024.

-This release was originally published on the NC3RS website and has been edited for style.