EUDOC on the IBM Blue Gene/L system: Accelerating the transfer of drug discoveries from laboratory to patient

Abstract

EUDOCe is a molecular docking program that has successfully helped to identify new drug leads. This virtual screening (VS) tool identifies drug candidates by computationally testing the binding of these drugs to biologically important protein targets. This approach can reduce the research time required of biochemists, accelerating the identification of therapeutically useful drugs and helping to transfer discoveries from the laboratory to the patient. Migration of the EUDOC application code to the IBM Blue Gene/Le (BG/L) supercomputer has been highly successful. This migration led to a 200-fold improvement in elapsed time for a representative VS application benchmark. Three focus areas provided benefits. First, we enhanced the performance of serial code through application redesign, hand-tuning, and increased usage of SIMD (single-instruction, multiple-data) floating-point unit operations. Second, we studied computational load-balancing schemes to maximize processor utilization and application scalability for the massively parallel architecture of the BG/L system. Third, we greatly enhanced system I/O interaction design. We also identified and resolved severe performance bottlenecks, allowing for efficient performance on more than 4,000 processors. This paper describes specific improvements in each of the areas of focus.

Normal-Mode-Analysis–Monitored Energy Minimization Procedure for Generating Small–Molecule Bound Conformations

“…These findings suggest that small molecules prefer to adopt local minimum conformations when binding to proteins and theoretically support the virtual screening methods that use local minimum conformations to enable massively parallel docking [6], [8], [14]. In practice, the folding of small molecules caused by energy minimization in the absence of their partners hampers the generation of small–molecule bound conformations from their two–dimensional (2D) structures.…”

Preference of Small Molecules for Local Minimum Conformations when Binding to Proteins

“…The resulting local minimum conformations were then subject to a cluster analysis with consideration of molecular symmetry to remove duplicated or similar conformations. Different local minimum conformations of the ligand were then docked into the binding site of the protein structure taken from the complex crystal structure using the EUDOC program [6], [8], [24]. The EUDOC-generated protein-bound ligand complex with the strongest intermolecular interaction energy was compared to the corresponding complex crystal structure.…”

“…The conformation selection theory is ideal to computationally account for molecular flexibility in docking, because it can convert a ligand–receptor association best described by the conformational induction theory to a series of associations each of which can be described by the lock-key theory [6]. The conformation selection theory thereby affords parallel computing and enables a docking study to be performed on thousands of IBM Blue Gene processors with high processor utilization [6]–[8].…”

Structure-based discovery of dengue virus protease inhibitors

“…First, due to concerns regarding cellular uptake and cell membrane impermeability to ions (Ghose et al, 2001, Alberts et al, 2002) only neutral non-zwitterionic compounds (at pH 7.4) were passed to the filtered chemical library Second, only compounds readily purchasable from highly reputable chemical vendors were retained in the filtered chemical library. Using the EUDOC program (Pang et al, 2001 and 2008; Wang and Pang, 2007), we computationally screened this filtered chemical library against two previously reported dengue virus type 2 (DEN2V) protease crystal structures, the NS3 protease domain alone (PDB identifier 1BEF; Murthy et al, 1999) and the NS3 protease domain complexed with the Bowman-Birk inhibitor (a soybean protein that has been shown to inhibit the dengue protease) (PDB identifier 1DF9; Murthy et al, 2000). The EUDOC program performed virtual screens by systematically and separately docking each small molecule from our chemical library into the protease active site and the P1 pocket.…”