中国人民解放军陆军军医大学（第三军医大学）：研究生《生命科学综合》核心课程PPT教学课件_06 蛋白质结构与功能预测 Computational Methods in Proteins

Computational Methods in Proteins sep,162015 邹法云 Associate professor College of Basic Medical Sciences E-mail:lingyunzoul@gmail.com

Computational Methods in Proteins Associate Professor College of Basic Medical Sciences E-mail: lingyun.zou@gmail.com Sep, 16 2015

推荐参考书 Second edition 编著者 STRUCTURAL BIOINFORMATICS Jenny gu, philip e. bourne 才出版商 John wiley sons lenny Gu, PhD 2009年3月16日 Philip E. Bourne, PhD AyWILEY-BLACKWELL

推荐参考书 编著者： Jenny Gu, Philip E. Bourne 出版商： John Wiley & Sons 2009年3月16日

Outline Protein Structure Prediction Alignment (蛋白质结构预测&比对) Molecular Dynamics simulation (分子动力学模拟) Protein-protein interaction prediction (蛋白质蛋白质相互作用预测)

Outline • Protein Structure Prediction & Alignment (蛋白质结构预测&比对) • Molecular Dynamics Simulation (分子动力学模拟) • Protein-protein interaction prediction (蛋白质-蛋白质相互作用预测)

Protein Structure Prediction Alignment Molecular Dynamics Simulation Protein-protein interaction prediction

• Protein Structure Prediction & Alignment • Molecular Dynamics Simulation • Protein-protein interaction prediction

蛋白质结构的实验测定 Xray晶体衍射 需要生长蛋白质晶体(这对一部分蛋白质几乎是不 可能的,总之,不容易) 衍射图样能进行反傅立叶变换来表征电子密度(这 有“相”的问题) 核磁共振谱(NMR) 需要高纯度的样本 只适用于相对较小的蛋白质

蛋白质结构的实验测定 ⚫ X-ray 晶体衍射 – 需要生长蛋白质晶体（这对一部分蛋白质几乎是不 可能的，总之，不容易） – 衍射图样能进行反傅立叶变换来表征电子密度（这 有“相”的问题） ⚫ 核磁共振谱（NMR） – 需要高纯度的样本 – 只适用于相对较小的蛋白质

X-Ray Crystallography X-ray Detector Diffraction X-rays Computer Graphics Protein Model Electron Density Map

• X-Ray Crystallography

什么是蛋白质结构预测? Predict the 3-dimensional structure of a protein from its primary sequence GICPRFAHVIENLLLGTPS SYETSLKEFEPDDTMKDA GMQMKKVLDSLPQTTRE NIMKLTEKⅣ KSPLCM 生物信息学难题之

什么是蛋白质结构预测？ GICPRFAHVIENLLLGTPS SYETSLKEFEPDDTMKDA GMQMKKVLDSLPQTTRE NIMKLTEKIVKSPLCM ！ • Predict the 3-dimensional structure of a protein from its primary sequence 生物信息学难题之一

蛋白质结构预测方法 Homology modeling(同源建模) typically >20-30% identity Fold recognition(折叠识别) cannot detect similarity in sequence level Ab initio( de novo) modeling(从头计算) v cannot detect fold

蛋白质结构预测方法 • Homology modeling （同源建模） ✓ typically >20-30% identity • Fold recognition（折叠识别） ✓ cannot detect similarity in sequence level • Ab initio (de novo) modeling（从头计算） ✓ cannot detect fold

脱氢酶/还原酶家族成员7B Homology model of the dhrs7B protein created with SwissModel and rendered with Pymol

Homology model of the DHRS7B protein created with SwissModel and rendered with PyMOL. 脱氢酶/还原酶家族成员7B

Why use homology modeling? ° Ab Initio protein folding(“ randon” sampling) Y 100 aa, 3 conf /residue gives approximately 1048 different overall conformations Random sampling is NoT feasible, even if conformations can be sampled at picosecond (10-2 sec)rates. Do homology modelling instead

Why use homology modeling? • Ab Initio protein folding (“random” sampling): ✓ 100 aa, 3 conf./residue gives approximately 1048 different overall conformations! • Random sampling is NOT feasible, even if conformations can be sampled at picosecond (10-12 sec) rates. • Do homology modelling instead