Hello there,

I am trying to build a ligand-based pharmacophore model on Ligan Scout 4.4. I chose a training set of 18 highly to moderately active compounds from a list of 77 compounds that share a common scaffold. I generated conformers for the entire dataset, and performed clustering to get an understanding of the similarity between compounds. For the test set, I included a range of 6 molecules from highly active to inactive. For the inactive compounds, I unchecked the activity box. The remaining compounds were ignored, and the whole dataset of training and test datasets was composed of 24 molecules.

My issue is when I try to generate a merged features pharmacophore model using the pharmacophore fit and atom overlap, the program loads, and at the end there's no pharmacophore output. There isn't even an error message? Is this normal? Does this mean there was no high enough fit score? Because I did the same with much less structurally similar compounds and got several models.