• Korean Journal of Veterinary Service
• /
• v.42 no.2
• /
• pp.73-83
• /
• 2019

Foot-and-Mouth Disease (FMD) is a highly contagious trans-boundary viral disease caused by FMD virus, which causes huge economic losses. FMDV infects cloven hoofed (two-toed) mammals such as cattle, sheep, goats, pigs and various wildlife species. To control the FMDV, it is necessary to understand the life cycle and the pathogenesis of FMDV in host. Especially, the protein-protein interaction between FMDV and host will help to understand the survival cycle of viruses in host cell and establish new therapeutic strategies. However, the computational approach for protein-protein interaction between FMDV and pig hosts have not been applied to studies of the onset mechanism of FMDV. In the present work, we have performed the prediction of the pig's proteins which interact with FMDV based on RNA-Seq data, protein sequence, and structure information. After identifying the virus-host interaction, we looked for meaningful pathways and anticipated changes in the host caused by infection with FMDV. A total of 78 proteins of pig were predicted as interacting with FMDV. The 156 interactions include 94 interactions predicted by sequence-based method and the 62 interactions predicted by structure-based method using domain information. The protein interaction network contained integrin as well as STYK1, VTCN1, IDO1, CDH3, SLA-DQB1, FER, and FGFR2 which were related to the up-regulation of inflammation and the down-regulation of cell adhesion and host defense systems such as macrophage and leukocytes. These results provide clues to the knowledge and mechanism of how FMDV affects the host cell.

• Proceedings of the Korean Biophysical Society Conference
• /
• 2002.06b
• /
• pp.13-13
• /
• 2002

기계학습(maching learning)은 경험을 통한 테이터 관측으로부터 스스로 성능을 향상할 수 있는 컴퓨터를 연구하는 인공지능(artificial intelligence)의 한 연구분야이다. 최근 들어 기계학습은 Bioinformatics 분야에서 생명과학 데이터마이닝을 위한 하나의 핵심기술로 부상하고 있다.(중략)

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.12
• /
• pp.3011-3016
• /
• 2015

Representing protein three-dimensional structure by concatenating a sequence of protein fragments gives an efficient application in analysis, modeling, search, and prediction of protein structures. This paper investigated the effective combination of distance measures, which can exploit large protein structure database, in order to construct a protein fragment library representing native protein structures accurately. Clustering method was used to construct a protein fragment library. Initial clustering stage used inter alpha-carbon distance having low time complexity, and cluster extension stage used the combination of inter alpha-carbon distance, Binet-Cauchy distance, and root mean square deviation. Protein fragment library was constructed by leveraging large protein structure database using the proposed combination of distance measures. This library gives low root mean square deviation in the experiments representing protein structures with protein fragments.

• Proceeding of EDISON Challenge
• /
• 2014.03a
• /
• pp.63-74
• /
• 2014

Predicting protein loop structures is an important modeling problem since protein loops are often involved in diverse biological functions by participating in enzyme active sites, ligand binding sites, etc. However, loop structure prediction is difficult even when structures of homologous proteins are known due to large sequence and structure variability among loops of homologous proteins. Therefore, an ab initio approach is necessary to solve loop modeling problems. One of the difficulties in the development of ab initio loop modeling method is to derive an accurate scoring function that closely approximates the true free energy function. In particular, entropy as well as energy contribution have to be considered adequately for loops because loops tend to be flexible compared to other parts of protein. In this study, the contribution of conformational entropy is considered in scoring loop conformations by employing "colony energy" which was previously proposed to estimate the free energy for an ensemble of conformations. Loop conformations were generated by using two EDISON_Chem programs GalaxyFill and GalaxySC, and colony energy was designed for this sampling by tuning relevant parameters. On a test set of 40 loops, the accuracy of predicted loop structure improved on average by scoring with the colony energy compared to scoring by energy alone. In addition, high correlation between colony energy and deviation from the native structure suggested that more extensive sampling can further improve the prediction accuracy. In another test on 6 ligand-binding loops that show conformational changes by ligand binding, both ligand-free and ligand-bound states could be identified by using colony energy when no information on the ligand-bound conformation is used.

• Journal of Integrative Natural Science
• /
• v.9 no.1
• /
• pp.35-40
• /
• 2016

Protein phosphatase manganese dependent 1D (PPM1D) is one of the Ser/Thr protein phosphatases belongs to the PP2C family. They play an important role in cancer tumorigenesis of various tumors including neuroblastoma, pancreatic adenocarcinoma, medulloblastoma, breast cancer, prostate cancer and ovarian cancer. Even though PPM1D is involved in the pathophysiology of various tumors, the three dimensional protein structure is still unknown. Hence in the present study, homology modelling of PPM1D was performed. 20 different models were modelled using single- and multiple-template based homology modelling and validated using different techniques. Best models were selected based on the validation. Three models were selected and found to have similar structures. The predicted models may be useful as a tool in studying the pathophysiological role of PPM1D.

• Bioinformatics and Biosystems
• /
• v.1 no.1
• /
• pp.82-87
• /
• 2006

The most popular protein structure prediction method is comparative modeling. To guarantee accurate comparative modeling, the sequence alignment between a query protein and a template should be accurate. Although choosing the best template based on the protein sequence alignments is most critical to perform more accurate fold-recognition in comparative modeling, even more critical is the sequence alignment quality. Contrast to a lot of attention to developing a method for choosing the best template, prediction of alignment accuracy has not gained much interest. Here, we develop a method for prediction of the shift score, a recently proposed measure for alignment quality. We apply support vector regression (SVR) to predict shift score. The alignment between a query protein and a template protein of length n in our own library is transformed into an input vector of length n +2. Structural alignments are assumed to be the best alignment, and SVR is trained to predict the shift score between structural alignment and profile-profile alignment of a query protein to a template protein. The performance is assessed by Pearson correlation coefficient. The trained SVR predicts shift score with the correlation between observed and predicted shift score of 0.80.

• 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.

• Korean Journal of Microbiology
• /
• v.39 no.2
• /
• pp.67-75
• /
• 2003

One large topic in comparative genomics is to predict functional annotation by classifying protein sequences. Computational approaches for function prediction include protein structure prediction, sequence alignment and domain prediction or binding site prediction. This paper is on another computational approach searching for sets of homologous sequences from sequence similarity graph. Methods based on similarity graph do not need previous knowledges about sequences, but largely depend on the researcher's subjective threshold settings. In this paper, we propose a genome sequence clustering method of iterative testing and graph decomposition, and a simple method to calculate a strict threshold having biochemical meaning. Proposed method was applied to known bacterial genome sequences and the result was shown with the BAG algorithm's. Result clusters are lacking some completeness, but the confidence level is very high and the method does not need user-defined thresholds.

• Journal of the Korean Magnetic Resonance Society
• /
• v.15 no.1
• /
• pp.69-79
• /
• 2011

HP0827 has two RNP motif which is a very common protein domain involved in recognition of a wide range of ssRNA/DNA.We acquired 3D NMR spectra of HP0827 which shows well dispersed and homogeneous signals which allows us to assign 98% of all $^1H_N$, $^{15}N$, $^{13}C_{\alpha}$, $^{13}C_{\beta}$ and $^{13}C$=O resonances and 90% of all sidechain resonances. The sequence-specific backbone resonance assignment of HP0827 can be used to gain deeper insights into the nucleic acids binding specificity of HP0827 in the future study. Here, we report secondary structure prediction of HP0827 derived from NMR data. Additionally, ssRNA/DNA binding assay studies was also conducted. This study might provide a clue for exact function of HP0827 based on structure and sequence.

• The KIPS Transactions:PartD
• /
• v.19D no.1
• /
• pp.21-28
• /
• 2012

Predication of protein interaction sites for monomer structures can reduce the search space for protein docking and has been regarded as very significant for predicting unknown functions of proteins from their interacting proteins whose functions are known. In the other hand, the prediction of interaction sites has been limited in crystallizing weakly interacting complexes which are transient and do not form the complexes stable enough for obtaining experimental structures by crystallization or even NMR for the most important protein-protein interactions. This work reports the calculation of 3D surface patches of complex structures and their properties and a machine learning approach to build a predictive model for the 3D surface patches in interaction and non-interaction sites using support vector machine. To overcome classification problems for class imbalanced data, we employed an under-sampling technique. 9 properties of the patches were calculated from amino acid compositions and secondary structure elements. With 10 fold cross validation, the predictive model built from SVM achieved an accuracy of 92.7% for classification of 3D patches in interaction and non-interaction sites from 147 complexes.