Predict absorption
Draw a structure, or paste a SMILES and it will be drawn. The answer comes back as a bar with a shaded interval, because that is the width the data supports.
Each one loads into the editor and predicts itself.
The absorption model is fitted to small drug-like molecules. A cyclic peptide belongs on the peptide page, where two assay arms are reported separately and never pooled.
How to read the bar
The shaded band is the interval, set by the root mean square error the model made on compounds it never saw. Held-out error is 21.3 plus or minus 1.3 percent absorption units across 12 independent stratified splits, and the deployed committee scores 20.9. Held-out correlation between predicted and measured is 0.744. A predictor that answers with the training mean on the same rows scores about 29 to 31, which is the floor rather than a result: any honest model has to beat it, and beating it by ten units is what this one does.
The bar is useful for telling a well absorbed compound from a poorly absorbed one and for ranking a series against itself. It is not useful for telling 85 apart from 90. If two compounds in a series have overlapping bands, the model has not separated them.
A prediction from a sparse or distant region of the training set is flagged in a box above the result, not in a footnote. A flag is an answer: it says the model has nothing reliable to offer for this structure. It is never hidden and never quietly filled in. The full method is on How it works and the training data on Data.