Chemical similarity searches using latent semantic structural indexing (LaSSI) and comparison to TOPOSIM

Similarity searches based on chemical descriptors have proven extremely use ful in aiding large-scale drug screening. Here we present results of simila rity searching using Latent Semantic Structure Indexing (LaSSI). LaSSI uses a singular value decomposition on chemical descriptors to project molecule s into a h-dimensional descriptor space, where h is the number of retained singular values. The effect of the projection is that certain descriptors a re emphasized over others and some descriptors may count as partially equiv alent to others. We compare LaSSI searches to searches done with TOPOSIM, o ur standard in-house method, which uses the Dice similarity definition. Sta ndard descriptor-based methods such as TOPOSIM count all descriptors equall y and treat all descriptors as independent. For this work we use atom pairs and topological torsions as examples of chemical descriptors. Using object ive criteria to determine how effective one similarity method is versus ano ther in selecting active compounds from a large database, we find for a ser ies of 16 drug-like probes that LaSSI is as good as or better than TOPOSIM in selecting active compounds from the MDDR database, if the user is allowe d to treat h as an adjustable parameter. Typically, LaSSI selects very diff erent sets of actives than does TOPOSIM, so it can find classes of actives that TOPOSIM would miss.