Protein preparation is an integral part of structure-based drug design and many steps are involved in the process. In this presentation, we will explore the various steps of preparing proteins and the implications of each step in virtual screening. We have systematically examined a broad range of factors, including Asn/Gln/His flips, hydrogen bond optimization, all-atom refinement, ligand tautomer/ionization state, and the treatment of crystallographic water molecules and have found that the various factors have different levels of importance depending on the system. We present a complete preparation protocol using the Protein Preparation Wizard, which results in significantly improved enrichments over preparations with only a subset of the parameters varied.

Canvas is a feature-rich cheminformatics package, and two highly-anticipated additional applications have been implemented over the last year. In this presentation, Senior Scientist Yuling An will discuss the Canvas Hole Filling (HF) method for library optimization as well as kernel-based PLS (KPLS) for building QSAR models.

Canvas HF is an automated method to fill holes in an existing library using compounds from an external source. The method uses distances computed from 2D chemical fingerprints and selects compounds that fill vacuous regions while not suffering from the problem of selecting only compounds at the edge of the chemical space. Dr. An will also present an extension of the method in which chemical properties can be considered with the selection process to bias the compounds toward a desired property space.

The KPLS algorithm, which can operate on molecular fingerprints, builds robust models for explaining and predicting SAR, with prediction uncertainties calculated on a large test set showing a positive relationship with the absolute errors. Once a model is built, the atomic contributions of model sensitivities can be overlaid on 2D molecular representations, allowing visual analysis.

The most interesting molecular events happen over micro and millisecond time scales that are hard to reach using conventional molecular dynamic simulations. Kinases in particular are highly flexible and undergo large conformational changes that can affect both substrate and inhibitor binding. We use a novel type of enhanced sampling method called metadynamics to explore motions and movements in kinases. In this presentation, we will discuss a specific example around a Parkinson's disease linked mutation G2019S in LRRK2 kinase. Using metadynamics, we are able to explain the unusual binding mode of inhibitors to the mutated kinase and how the insights from the metadynamics simulation can be used to design inhibitors for this kinase in the future.

Schrödinger's Vice President of Applications Science presents an overview of scientific advances, along with new features and enhancements in Schrödinger Suite 2012. Among the topics to be discussed are:

• Native Mac support

• Improved Maestro graphics

• Improved Ligand Interaction Diagram (LID)

• Running KNIME workflows directly from Maestro

• Running Schrödinger software on the Amazon Cloud

• SiteMap detection of shallow binding sites for
   protein-protein interactions

• Enhanced Maestro productivity tools

The Schrödinger Cloud Package delivers the Schrödinger Suite on an easy-to-use virtual cluster in the Amazon cloud. A fully managed infrastructure, use-based pricing, and on-demand resources are just some aspects of the Schrödinger Cloud Package that make it an attractive alternative to dedicated, in-house resources. This seminar will focus on these, and other technical details of the Schrödinger Cloud Package.