초록문의 abstract@kcsnet.or.kr

결제문의 member@kcsnet.or.kr

현재 가능한 작업은 아래와 같습니다.
  • 02월 20일 17시 이후 : 초록수정 불가능, 일정확인 및 검색만 가능

제113회 대한화학회 학술발표회, 총회 및 기기전시회 안내 Dockable Pocket Site Prediction (DPSP) for annotation of protein function

2014년 2월 13일 16시 45분 46초
PHYS.P-417 이곳을 클릭하시면 발표코드에 대한 설명을 보실 수 있습니다.
4월 16일 (수요일) 16:00~19:00
저자 및
임규태, 김봉주1, 이진혁2,*
과학기술연합대학원대학교(UST) 생명정보학과, Korea
1동아대학교 생명과학과, Korea
2한국생명공학연구원 생명정보센터, Korea
Protein binding pocket is essential for signal transduction and reactant generation in vivo, so information of ligand and protein binding site position is important for understanding mechanisms. However, most of protein-ligand binding site and interaction details are unknown. Here, DPSP, a method to find protein binding site using geometry method and annotate ligand information based on protein only structure database. DPSP finds match protein sequence in database which has 84,600 structures and ligand information is gathered. After finding matched protein structures, DPSP chooses binding site which has high Matthews correlation coefficient and implements ligand docking process for more specific ligand position. DPSP was tested on 10 proteins which are CASP10 FN targets. In CASP10, we got a Matthews correlation coefficient about 0.27 among 10 targets. The top three scores have over 0.6 (30%). When we applied docking system, Matthews correlation coefficients were 0.25 (the lowest energy dose) and 0.49 (the best docking dose). This method is expected to use drug research efficiently in the way that it predicts binding site and ligand information of unknown protein binding mechanisms. Also, by providing various ligand candidates, it will help a researcher to select the ligand in researcher’s case.