You could build a heuristic risk score against each molecule:
- What functional groups does it have?
- How many functional groups does it have?
- How much electron delocalization does it have?
- How much of that electron delocalization is PAHs?
- Does the molecule participate in redox reactions?
Etc.
Basically check to see which molecules can generate free radicals, strip DNA, convert to dangerous metabolites, etc.
Once you have that trained, you'll be able to publish it and dramatically improve the current SoTA!
I wish it was as easy as this - while there are known toxicophores/no-go functional groups in medchem, there is going to be a big dearth of data on non-acute (e.g. not hERG, hepatotoxicity, etc) toxicity, which is really what the question is here: what are the marginal risks/rewards from eating existing food X (since we know it's probably not acutely toxic).
It's call QSAR https://en.wikipedia.org/wiki/Quantitative_structure%E2%80%9... They may use a hundred of properties to guess the effect of the molecule.