Computational methods have become an indispensable part of lead identification efforts. Nearly all methods require accurate 3D molecular models as a starting point. However, many corporate and purchasable compound databases contain only 2D molecular structures. Efficient and accurate 2D to 3D conversion is therefore a key precursor to computational analyses.

Beyond simple one-to-one structural conversion, it is equally important to generate scientifically sound molecular models that enumerate the different structural and chemical possibilities a ligand could sample, as these variations could lead to dramatically different results in subsequent computations. A versatile conversion program that can be configured to generate ligand libraries with the desired structural and chemical features can significantly streamline the entire in silico drug discovery process.