Using least median of squares for structural superposition of flexible proteins
Yu-Shen Liu1 Yi Fang1 Karthik Ramani1,2 |
|
Abstract |
The conventional superposition methods use an ordinary least
squares (LS) fit for structural comparison of two different
conformations of the same protein. The main problem of the LS fit
is very unstable and is sensitive to local displacements. To
overcome this problem, we present a new algorithm to overlap two
protein conformations by their atomic coordinates using a robust
statistics technique: least median of squares (LMS). In order to
effectively approximate the LMS optimization, the forward search
technique is utilized. Our algorithm can automatically detect and
superimpose the rigid core regions of two conformations with small
or large displacements. In contrast, most existing superposition
techniques, which strongly depend on the initial LS estimating for
the entire atom sets of proteins, usually fail on structural
superposition of two conformations with large displacements. The
proposed algorithm is robust and does not require any prior
knowledge of the flexible regions. Our fit tool has produced
successful superpositions when applied to proteins for which two
conformations are known. |
Downloads
|
|
Examples
|
Multiple level superposition for Topo II: 1bgw (red) and 1bjt (green). (a) Level 1 (Core% = 56.4%). (b) Level 2 (Core% = 22.1%). (c) Level 3 (Core% = 11.7%). (d) Level 4 (Core% = 5.1%). Note that our method can capture different rigid domains in multiple level superposition, where the superimposed rigid domains are highlighted in the selected regions with the solid line boundary.
|
Acknowledgment |
We would like to thank Dr. Talapady Bhat for some helpful comments
during our work. This material is partly based upon work supported
by the National Science Foundation under Grant IIS No. 0535156. |