Biomedical subjects
Publications and source records attributed to Vicky Choi.
In this paper, we present a new algorithm, which is based on an efficient heuristic for local search, for rigid protein-small-molecule docking. We tested our algorithm, called Yucca, on the recent 100-complex benchmark, using the conformer generator OMEGA to generate a set of low-energy conformers. The results showed that Yucca is competitive both in terms of algorithm efficiency and docking accuracy.
We propose an assembly algorithm Barnacle for sequences generated by the clone-based approach. We illustrate our approach by assembling the human genome. Our novel method abandons the original physical-mapping-first framework. As we show, Barnacle more effectively resolves conflicts due to repeated sequences which is the main difficulty of the sequence assembly problem. In addition, we are able to detect inconsistencies in the underlying data. We present and compare our results on the December 2001 freeze of the public working draft of the human genome with NCBI's assembly (Build 28). The assembly of December 2001 freeze of the public working draft generated by Barnacle and the source code of Barnacle are available at (http://www.cs.rutgers.edu/~vchoi).
A conformation of a molecule is defined by the relative positions of atoms and by the chirality of asymmetric atoms in the molecule. The three main representations for conformations of molecules are Cartesian coordinates, a distance geometry descriptor (which consists of a distance matrix and the signs of the volumes of quadruples of atoms), and internal coordinates. In biochemistry, conformational changes of a molecule are usually described in terms of internal coordinates. However, for many applications, such as molecular docking, the Cartesian coordinates of atoms are needed for computation. Although, for each conformational change, the Cartesian coordinates of atoms can be updated in linear time (which is optimal asymptotically), the constant factor becomes significant if a large number of updates are needed. Zhang and Kavraki (J. Chem. Inf. Comput. Sci. 2002, 42, 64-70) examined three methods: the simple rotations, the Denavit-Hartenberg local frames, and the atom-group local frames. On the basis of their implementations, they showed that the atom-group local frames are more efficient than the other two. In this paper, by expressing the torsion-angle change as a composition of translations and rotations, we observe that the simple rotations can be implemented in an efficient way by taking advantage of consecutive operations. Both quantitative and experimental comparisons show that the improved simple rotations, in which rotations are expressed in unit quaternions, are as efficient as the atom-group local frames and, thus, have the advantage of avoiding the need of precomputations of a set of local frames and transformations between them.