3차원 형상 분석 기반 단백질 유사성 분석
Structural Similarity Evaluation between Proteins Based on 3D Model Shape Analysis
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
단백질 구조 분석 및 비교는 X-ray 결정 방식이나 NMR 분광기 분석과 같은 실험적인 방법으로 수세기 동안 진행되어 십삼만개 이상의 원자 수준에서의 고해상도 단백질 3차 구조를 파악하고 있다. X-ray 결정 방식은 결정(Crystal structure)을 만들 수 없는 단백질은 구조 분석이 안 되는 단점이 있고, NMR 분광기 분석은 단백질 결합 등의 이유로 실험대상의 크기가 커지면 분석을 못하는 한계가 있다. 하지만 단백질의약품의 구조적 유사성을 평가하는 것은 단백질의약품 개발과 생산과정에서 아주 중요한 부분이다. 현재 이러한 유사성 평가는 다양한 분광학적 방법을 활용하지만 기술적 제한으로 인해 구조적인 정보를 확인하기에는 한계가 있다. 더구나 최근 Single Angle X-ray Scattering(SAXS) 분석법을 통해 해상도가 낮은 표면 구조 정보를 보다 쉽게 확보할 수 있게 되었다. 이에 본 연구에서는 저비용/단시간에 확보할 수 있는 저해상도 3차 구조 (표면 구조) 정보를 활용하여 동등성을 평가할 수 있는 기술을 개발하고자 하였다. 본 연구에서는 단백질 저해상도 모델로 표면구조 데이터의 특성을 이용하여 3D 모델 형태 특성 데이터를 기반으로 한 새로운 형태 기술자(Shape Descriptor)를 정의하고 이를 활용하여 단백질의 구조적 유사성을 평가할 수 있는 방법을 제안한다. 제안한 방법은 기존 방법인 고 해상도 단백질 3차 구조 정보를 활용하여 유사성을 판별한 RMSD 값과 비교하여 제안된 형태 기술자를 적용한 방법의 실효성을 검증하였다.
Abstract
Over 130,000 protein structures are solved and deposited in the PDB with molecular structural models of high resolution determined by X-ray diffraction or NMR studies. Those conventional methods for structural determination and similarity comparison are not applicable for therapeutic proteins due to their own technical limitations. However, assessing structural comparability for therapeutic proteins is critical during biopharmaceutical development or manufacturing processes. Currently, those assessments were done by various spectroscopic methods but those methods are not giving specific structural information. Moreover, recent developments such as single angle X-ray scattering (SAXS) techniques can provide low-resolution structural models for proteins more easily than those for X-ray diffraction or NMR studies. These enable us to develop fast and reliable method for the similarity assessment of low-resolution surface models of therapeutic proteins using geometric shape descriptor. Our descriptor consists of two features, local and global. The local feature calculates the distance distribution for each vertex from the center in 510 bins. For global feature, the ratio, x to y axis and x to z axis from the surface model of proteins were determined. A geometric shape descriptor is then constructed by combining local and global features with weights. We applied this geometric shape descriptor to assess structural similarity for the therapeutic protein, insulin models. Our geometric shape descriptor can clearly classify human insulin models and insulin analog models which having locally different structures. The performance of the geometric shape descriptor was evaluated by comparing to the conventional method of the root mean square deviation measure (RMSD) which computes the minimum average distance between the backbones of superimposed. The result shows that the performance of the global shape descriptor is comparable to that of RMSD. This provides potential applications to classify protein structure and compare low-resolution protein structures for EM and SAXS.
참고문헌 (19)
- A. Kryshtafovych, "Some of the most interesting CASP11 targets through the eyes of their authors", Proteins, 84, pp.34~50 (2016)
- M. F. Lensink, "Modeling protein–protein and protein–peptide complexe", pp.359~377 (2017), Proteins
- A. G. Kikhney, "A practical guide to small angle X-ray scattering (SAXS) of flexible and intrinsically disordered proteins", FEBS Letters, 589(19), pp.2570~2577 (2015)
- S. Mahmoudi, "3D models retrieval by using characteristic views", CPR’02, pp.11~15 (2002)
- J-L. Shih, "A new 3D model retrieval approach based on the elevation descriptor", Pattern Recognition, 40(1), pp.283~295 (2007)
- D-Y. Chen, "On visual similarity based 3d model retrieval", Proceedings of Eurographics, pp.223~232 (2003)
- R. Ohbuchi, "Salient local visual features for shape-based 3D model retrieval", Proceedings of the IEEE international conference on shape modeling and applications (SMI 2008), pp.93~102 (2008)
- D. Vranic, "Tools for 3d-object retrieval: Karhunen-loeve transform and spherical harmonics", Proceedings of the IEEE fourth workshop on multimedia signal processing, pp.293~229 (2001)
- R. Osada, "Matching 3D models with shape distributions", Proceedings of the IEEE international conference on shape modeling and applications (SMI2001), pp.154~166 (2001)
- R. Ohbuchi, "Shape similarity search of 3D models by using enhanced shape functions", International Journal of Computer Applications in Technology, 23(2), pp.97~104 (2005)
- Y. Liu, "The generalized shape distributions for shape matching and analysis", Proceedings of the IEEE international conference on shape modeling and applications (SMI2006) (2006)
- M. Ankerst, "3D shape histograms for similarity search and classification in spatial databases", Proceedings of the 6th international symposium of spatial databases (SSD1999) (1999)
- M. Hilaga, "Topology matching for fully automatic similarity estimation of 3D shapes", Proceedings of ACM SIGGRAPH 2001, pp.203~212 (2001)
- D. Duhovny, "Efficient unbound docking of rigid molecules", Proceedings of the Second International Workshop on Algorithms in Bioinformatics, pp.185~200 (2002), SpringerVerlag
- D. Schneidman-Duhovny, "Geometry-based flexible and symmetric protein docking", Proteins, 60(2), pp.224~231 (2005)
- V. Venkatraman, "Protein-protein docking using region-based 3D Zernike descriptors", BMC Bioinformatics, 10(407), pp.1~21 (2009)
- D. Kihara, "Molecular surface representation using 3D Zernike descriptors for protein shape comparison and docking", Current Protein & Peptide Science, 12(6), pp.520~530 (2011)
- Z. Shentu, "Context shapes: efficient complementary shape matching for protein–protein docking", Proteins, 70, pp.1056~1073 (2008)
- M. P. Favero-Retto, "Structural meta-analysis of regular human insulin in pharmaceutical formulations", European Journal of Pharmaceutics and Biopharmaceutics, 85, pp.1112~1121 (2013)