• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.22 no.1
• /
• pp.26-32
• /
• 2018

The protein secondary structures are important information for studying the evolution, structure and function of proteins. Recently, deep learning methods have been actively applied to predict the secondary structure of proteins using only protein sequence information. In these methods, widely used input features are protein profiles transformed from protein sequences. In this paper, to obtain an effective protein profiles, protein profiles were constructed using protein sequence search methods such as PSI-BLAST and HHblits. We adjust the similarity threshold for determining the homologous protein sequence used in constructing the protein profile and the number of iterations of the profile construction using the homologous sequence information. We used the protein profiles as inputs to convolutional neural networks and recurrent neural networks to predict the secondary structures. The protein profile that was created by adding evolutionary information only once was effective.

• BMB Reports
• /
• v.29 no.4
• /
• pp.380-385
• /
• 1996

Ovomucoid third domain is a serine proteinase inhibitor protein which consists of 56 amino acid residues. A fifty picosecond molecular dynamics (MD) simulation was carried out for ovomucoid third domain protein with 5 $\AA$ layer of water molecules. A comparison of main chain atoms in the MD averaged structure with the crystal structure showed that most of the backbone structures are maintained during the simulation. Investigation of the intramolecular hydrogen bondings indicated that most of the interactions between main chain atoms were conserved, whereas those between side chains were reorganized for the period of the simulation. Especially, the side chain interactions around the scissile bond of reactive site P1 (Met18) were found to be more extensive for the MD structures. During the simulation, hydrogen bonds were maintained between the side chains of Glu19 and Arg21 as well as those of Thr17 and Glu19. Extensive side chain interactions observed in the MD structures may shed light on the question of why protein proteinase inhibitors are strong inhibitors for proteinases rather than good substrates.

• Journal of Microbiology and Biotechnology
• /
• v.10 no.5
• /
• pp.733-736
• /
• 2000

Since pathogens resistant against vancomycin occur rapidly, the development of a new drug is needed. To make a new drug based on a rational drug design, the structural study of vancomycin is necessary. Accordingly, this study reports on a comparison of the solution structure of vancomycin determined by NMR spectroscopy, which was performed in the present work, with the X-ray crystallographic structure previously deposited in the Protein Data Bank (PDB).

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2005.09a
• /
• pp.256-259
• /
• 2005

Among non-synonymous SNPs that cause amino acid change in the protein product, the selection of disease-causing SNPs has been of great interest. We present the comparison between the evolutionary (SIFT score) and structural information (binding pocket) to show that the incorporation between them provides an advantage of sorting disease-causing SNPs from normal SNPs. To set up the procedure, we apply the machine learning method to the test data set from the laboratory experiments.

• Korean Chemical Engineering Research
• /
• v.46 no.3
• /
• pp.529-534
• /
• 2008

The normal shell and the regenerated oyster shell, Crassostrea gigas, are separated according to the characteristics of inner shell morphology. To study characteristics of chitin obtained from the regenerated shell, chitin prepared by acid and alkali process is analyzed by FT-IR (Fourier transform infrared spectrometer) and XRD (X-ray Diffractometer). The content of insoluble protein in the normal shell was more than doubled as compared with that in the regenerated shell. A comparison of secondary structure of the normal shell and the regenerated shell revealed that the content of random of the regenerated shell was above 47%, indicating an amount in the structural unordered state. Through amino acid composition analysis and secondary protein structure of soluble protein isolated from the normal shell and the regenerated shell, it was found that there are differences in biomineralization strategy of the regenerated shell as compared to the normal shell. The relatively low hardness of the regenerated shell is caused by the change of amino acid composition and ordered secondary protein structure as compared to hardness of the normal shell.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.10e
• /
• pp.613-615
• /
• 2002

단백질의 3차원 공간상의 구조는 단백질 기능을 파악하는데 중요한 정보를 제공하고 있다. 단백질간 구조 비교 방법은 기능적 또는 구조적으로 연관된 단백질 분류 및 단백질 모티프(motif)를 찾는데 유용하게 사용되고 있다. 본 논문에서는 단백질 이차 구조($\alpha$-나선구조와 $\beta$-병풍구조)와 그들 사이의 관계(각도, 거리, 길이, 수소결합)를 기반으로 표현된 두 단백질 구조에서 유사한 부분 구조를 찾는 방법에 대하여 기술한다. 제안된 단백질간 구조 사이의 유사한 부분구조를 찾는 방법은 두 단백질 구조론 이차 구조와 그들 사이의 관계를 이용하여 그래프를 형성한 후, 최대 유사 서브 그래프를 찾는 방법을 이용하여 유사한 부분구조를 찾을 수 있다.

• Asian Pacific Journal of Cancer Prevention
• /
• v.14 no.12
• /
• pp.7473-7482
• /
• 2013

The G-protein coupled receptor 87 (GPR87) is a recently discovered orphan GPCR which means that the search of their endogenous ligands has been a novel challenge. GPR87 has been shown to be overexpressed in squamous cell carcinomas (SCCs) or adenocarcinomas in lungs and bladder. The 3D structure of GPR87 was here modeled using two templates (2VT4 and 2ZIY) by a threading method. Functional assignment of GPR87 by SVM revealed that along with transporter activity, various novel functions were predicted. The 3D structure was further validated by comparison with structural features of the templates through Verify-3D, ProSA and ERRAT for determining correct stereochemical parameters. The resulting model was evaluated by Ramachandran plot and good 3D structure compatibility was evidenced by DOPE score. Molecular dynamics simulation and solvation of protein were studied through explicit spherical boundaries with a harmonic restraint membrane water system. A DRY-motif (Asp-Arg-Tyr sequence) was found at the end of transmembrane helix3, where GPCR binds and thus activation of signals is transduced. In a search for better inhibitors of GPR87, in silico modification of some substrate ligands was carried out to form polar interactions with Arg115 and Lys296. Thus, this study provides early insights into the structure of a major drug target for SCCs.

• Journal of Dairy Science and Biotechnology
• /
• v.21 no.2
• /
• pp.138-147
• /
• 2003

We here analyzed and proposed the structures of the N-linked sugar chains of PAS-7 from bovine milk fat globule membrane. The N-linked sugar chains were liberated from PAS-7 by hydrazinolysis and, after modifying the reducing ends with 2-aminopyridine (PA), were separated into one neutral (7N,55%) and two acidic (7M mono-, 43%; 7D, di-, 2%) sugar chain roups. The latter were converted into neutral groups (7MN and 7DN) by sialidase digestion. The structure of each of these PA-neutral sugar chains was determined by sugar analysis, sequential exoglycosidase digestion, partial acetolysis, and 1H-NMR spectroscopy. The results show that the 10 sugar chains were of the biantennary complex type with and without fucose. The structure of 7N2A one of the major sugar chains, was proposed as; [structure: see text] A structural comparison between PAS-6 and -7 indicated that although they shared the same protein core, their sugar moiety was markedly different, involving the existence of a different pathway during the post-transcriptional modification.

• Applied Biological Chemistry
• /
• v.41 no.8
• /
• pp.572-577
• /
• 1998

The solution conformation of the human insulin precursor, preproinsulin, is described in terms of NMR spectral data. NMR experiments were performed on preproinsulin, whose structure was compared with the NMR structure of native human insulin. Despite the presence of the C-peptide and/or the signal peptide, secondary structure analyses indicate that the native structures of the A and B chains are well conserved even in preproinsulin. The observed relative robustness of the native structure in precursor forms permits further protein engineering experiments where the C-peptide or N-terminal signal sequence can be altered for the purpose of increasing expression or purification yields when producing recombinant human insulin.

• Journal of the Korean Magnetic Resonance Society
• /
• v.19 no.2
• /
• pp.74-82
• /
• 2015

Ryanodine receptor (RyR) is one of the two major $Ca^{2+}$ channels in membranes of intracellular $Ca^{2+}$ stores and is found in sarcoplasmic reticulum (SR), endoplasmic reticulum (ER). RyR1 is also the major calmodulin-binding protein of sarcoplasmic reticulum membranes. Residues 4064-4210 in the RyR1 polypeptide chain has similar primary sequence with calmodulin (CaM) and was designated as CaM-like domain (CaMLD). When expressed as a recombinant peptide, CaMLD showed several CaM-like properties in previous studies. Still, previous studies of CaMLD were focused on protein-protein interactions rather than its own properties. Here, we studied the expression of CaMLD and its sub-domains corresponding to each lobe of CaM in Escherichia coli. CaMLD could be obtained only as inclusion body, and it was refolded using urea solubilization followed by dialysis. Using spectroscopic approaches, such as NMR, circular dichroism, and gel filtration experiment, we found that the refolded CaMLD exists as nonspecific aggregate, even though it has alpha helical secondary structure. In comparison, the first half of CaMLD (R4061-4141) could be obtained as natively soluble protein with thioredoxin fusion. After the removal of the fusion tag, it exhibited folded and helical properties as shown by NMR and circular dichroism experiments. Its oligomeric status was different from CaMLD, existing as dimeric form in solution. However, the second half of the protein could not be obtained as soluble protein regardless of fusion tag. Based on these results, we believe that CaMLD, although similar to CaM in sequence, has quite different physicochemical properties and that the second half of the protein renders it the aggregative properties.