I am trying to build QSPR models in KNIME using R. I want to use the molecular fingerprints as dependent variables. When I use it, "R Learner" doesn't recognise the fingerprint column and results in an error.
Can someone please help me in this aspect? Do I have to convert the "bit" format fingerprint into R-readable format? If so, how and which?
Have a look at the attached workflow and let me know if it helps at all. One branch shows how to use the native Bayes fingerprint learner and predictor, while the other relies on the Erl Wood fingerprint expander to create a set of regular int columns that you can use to train your model.