For the newest version of the Toolbox, is there a documented process to use the toolbox to determine if new chemical could be added to a pre-existing group of chemicals for read across assessment?
Put a more practical way, I have a file containing approx 30 chemicals that have been grouped and I would like to load into the toolbox (which I can do). I would then like to compare new chemicals to the group by Similarity to see if the new chemical can be added to the group. I would like to do all of this inside the toolbox if possible.
Can you guide me as to how this can be done?
If you highlight only your created database containing the 30 substances, no other stuctures or data will be returned or interrogated. Therefore the new structure will be compared only to your specific database.
I have completed a similar assessment with success.
You can do this by building a profiler.
There is some general guidance on building profilers here:
I suspect you will have to do this by defining the level of similarity to each of your 30 chemicals.
In addition I would recommend that you also define your similarity based on the profiling results which are common to all the members of the group.
You can then run your new profiler on any list of chemicals and find chemicals which potentially belong to your group.
Hope this helps
I tried the Profiler using the Similarity Boundry Option. When I try to upload the training set using my SMI file, I get a List Index out of Bounds error? My SMI file works in other areas of the toolbox so I think it is ok. What does the error mean.
I don't think you can load a list of chemicals into your custom profiler.
You will have to define the similarity rules against each and every chemical in your group and then logically group them, e.g.
>50% similarity with chemical 1 from your group AND >50% similarity with chemical 2 from your group AND ...
But once again, as I mentioned above, it might make more sense not to build your profiler based on mathematical structural similarity, but according to the commonalities in structure in your group as well as the results from the other profilers, e.g.
Aromatic (according to organic functional groups) AND Schiff base former according to Protein binding by OASIS AND...