I have a second problem I would like modify the RetroPath2.0-Mods-iQSAR-v2 such that molecules are represented by the ‘Morgan’ descriptor. I want to draw the curve showing the maximum property predictions obtained vs. iteration number. I must to set up the maximum
number of iteration to 50 as it takes time to run. Finally I would like to draw the molecules having the highest predicted value at the first iteration and the last iteration.
I have to use RetroPath2.0-Mods-iQSAR-v2 workflow (downloadable via this link): https://www.myexperiment.org/workflows/5029.html
To the Input Node it is necessary to have:
-Should one consider explicit Hs? No
-Number of iteration : 100
-Maximum molecular weight for compounds: 1,000
-Source file (.csv) is: " source.csv " . It is in this file https://www.myexperiment.org/files/1927.html
-Rules files (.csv) is : “rules-chemistry-switch.csv”. It is in this file https://www.myexperiment.org/files/1927.html
-Number of subsets for the tournament selection : 10
-Number of best structures to keep in each subset : 2
All boxes must be ticked.
To the Output Node it is necessary to put
-Path to the result folder : choose a folder
-Name of the global result file (will be written inside the result folder) “result.csv”.
The training set for CSV Reader is in this file https://www.myexperiment.org/files/1927.html. It’s is name : "setny-training-set "
It is an interesting workflow. The output of results can take up to 15 minutes.
I do not understand too much the sequences.
Thank you very much if you have an idea !