초록문의 abstract@kcsnet.or.kr

결제문의 member@kcsnet.or.kr

현재 가능한 작업은 아래와 같습니다.
  • 03월 02일 17시 이후 : 초록수정 불가능, 일정확인 및 검색만 가능

제109회 대한화학회 학술발표회, 총회 및 기기전시회 안내 Predicting protein loops in inaccurate model structures

2012년 2월 16일 15시 36분 18초
PHYS.P-465 이곳을 클릭하시면 발표코드에 대한 설명을 보실 수 있습니다.
4월 25일 (수요일) 18:00~21:00
저자 및
이규리, 석차옥
서울대학교 화학부, Korea
Protein loops are often involved in biological functions by contributing to binding sites or active sites. Therefore, high-accuracy loop structure prediction is crucial for understanding and regulating protein functions. A common way to evaluate the performance of protein loop modeling methods is to reconstruct the loop in the native structure after removing the loop region only. However, one frequently has to model loops in exact structural environment of low-resolution experimental structures or template-based models. We have developed a new loop modeling method that is less sensitive to inaccurate environment than existing methods but still retains high performance in native loop reconstruction, by employing a hybrid energy that combines physics-based and knowledge-based energy terms. In a native loop reconstruction test, median RMSD of 0.7 and 1.0 Å was obtained for 8- and 12-residue loops, respectively. A test on proteins with artificially perturbed side chain conformations that represent low-resolution crystal structures resulted in median RMSD of 1.1 and 1.8 Å for 8- and 12-residues loops, respectively. When compared to the best-performing previous method applied to the same set, the improvement is 1.1 and 0.5 Å. We have also introduced more realistic environmental inaccuracy including protein backbones by actual comparative modeling and obtained highly acceptable results. The overall performance is very encouraging considering that no optimization of the inaccurate environment was attempted, implying that the present loop modeling method can be applied to more realistic problems with much more efficiency than before.