The ChemAxon predictor is commercial software. You input a chemical structure, and it calculates various physical properties. But you can register for the online version for free, and you then get a limited number of free uses, which is what I did: https://chemicalize.com/ The documentation here explains a bit how the solubility prediction works: https://docs.chemaxon.com/display/docs/Solubility+Predictor You are exactly right that the structure of KL gives hints to solubility. How it works is it takes the chemical structure, breaks it down in to small parts, and calculates a combined solubility of the parts based on a database of the known solubilities of thousands of reference chemicals. Gory details in this article: Hou, T. J., et al. "ADME evaluation in drug discovery. 4. Prediction of aqueous solubility based on atom contribution approach." Journal of chemical information and computer sciences 44.1 (2004): 266-275. You can find the notation to use for the chemical structure at the PubChem site: https://pubchem.ncbi.nlm.nih.gov/ For example if you go to pub chem and search for kavain, you find the "smiles structure" which is a way of completely describing the molecular structure in a line of text: COC1=CC(=O)OC(C1)C=CC2=CC=CC=C2 and use the above text as input to the ChemAxon predictor.