I want to introduce the blog posted by Lhasa.
Conducting read-across assessments has been a manual process, presenting numerous challenges that complicate the accurate assessment of AI limits for nitrosamines. These challenges include:
Complexity of nitrosamine structures
Diverse chemical structures
Metabolic pathways
Data availability and quality
Limited data
Variability in data sources
Regulatory challenges
Regulatory acceptance
Computational and methodological issues
Lack of standardised protocols
Subjectivity and expert judgment
This is why Lhasa suggests the tools, 1. In silico tools and 2. Data sharing and collaboration. Please check the original.