Kwang Su Jung

QSE: A new 3-D solvent exposure measure for the analysis of protein structure†

Solvent exposure of amino acids measures how deep residues are buried in tertiary structure of proteins, and hence it provides important information for analyzing and predicting protein structure and functions. Existing methods of calculating solvent exposure such as accessible surface area, relative accessible surface area, residue depth, contact number, and half‐sphere exposure still have some limitations. In this article, we propose a novel solvent exposure measure named quadrant‐sphere exposure (QSE) based on eight quadrants derived from spherical neighborhood. The proposed measure forms a microenvironment around Cα atom as a sphere with a radius of 13 Å, and subdivides it into eight quadrants according to a rectangular coordinate system constructed based on geometric relationships of backbone atoms. The number of neighboring Cα atoms whose labels are the same is given as the QSE value of the center Cα atom at hand. As evidenced by histograms that show very different distributions for different structure configurations, the proposed measure captures local properties that are characteristic for a residues eight‐directional neighborhood within a sphere. Compared with other measures, QSE provides a different view of solvent exposure, and provides information that is specific for different tertiary structure. As the experimental results show, QSE measure can potentially be used in protein structure analysis and predictions.

A 9-state hidden Markov model using protein secondary structure information for protein fold recognition

In protein fold recognition, the main disadvantage of hidden Markov models (HMMs) is the employment of large-scale model architectures which require large data sets and high computational resources for training. Also, HMMs must consider sequential information about secondary structures of proteins, to improve prediction performance and reduce model parameters. Therefore, we propose a novel method for protein fold recognition based on a hidden Markov model, called a 9-state HMM. The method can (i) reduce the number of states using secondary structure information about proteins for each fold and (ii) recognize protein folds more accurately than other HMMs.

Automatic orthologous-protein-clustering from multiple complete-genomes by the best reciprocal BLAST hits

Though the number of completely sequenced genomes quickly grows in recent years, the methods to predict protein functions by homology from the genomes have not been used sufficiently. It has been a successful technique to construct an OPCs(Orthologous Protein Clusters) with the best reciprocal BLAST hits from multiple complete-genomes. But it takes time-consuming-processes to make the OPCs with manual work. We, here, propose an automatic method that clusters OPs(Orthologous Proteins) from multiple complete-genomes, which is, to be extended, based on INPARANOID which is an automatic program to detect OPs between two complete-genomes. We also prove all possible clustering mathematically.

RNA secondary structure prediction with simple pseudoknots based on dynamic programming

RNA molecules are sequences of nucleotides that serve as more than mere intermediaries between DNA and proteins, e.g. as catalytic molecules. The sequence of nucleotides of an RNA molecule constitutes its primary structure, and the pattern of pairing between nucleotides determines the secondary structure of an RNA. Computational prediction of RNA secondary structure is among the few structure prediction problems that can be solved satisfactory in polynomial time. Most work has been done to predict structures that do not contain pseudoknots. Pseudoknots have generally been excluded from the prediction of RNA secondary structures due to its difficulty in modelling. In this paper, we present a computation the maximum number of base pairs of an RNA sequence with simple pseudoknots. Our approach is based on pseudoknot technique proposed by Akutsu. We show that a structure with the maximum possible number of base pairs could be deduced by a improved Nussinovs trace-back procedure. In our approach we also considered wobble base pairings (G·U). We introduce an implementation of RNA secondary structure prediction with simple pseudoknots based on dynamic programming algorithm. To evaluate our method we use the 15 sequences with simple pseudoknots of variable size from 19 to 25 nucleotides. We get our experimental data set from PseudoBase. Our program predicts simple pseudoknots with correct or almost correct structure for 53% sequences.

A personalized biological data management system based on BSML

A mass of biological sequence and structural information have been produced in biological laboratories since the techniques to get the sequences of genomes or proteins have been improved in HGP (Human Genome Project). Unfortunately, there are scarcely software packages available to deal with the biological data in most of biological laboratories and they are just stored in file formats. An integrated management system of biological data is required to manage the sequence and annotation data taken from other open databases to improve the analysis of sequence data in biological laboratories. We therefore suggest a personalized system to edit, store and retrieve biological information, and convert the formats of sequence data, as well as to integrate and manage data.

Investigating the binding area of protein surface using MCL algorithm

Proteins combine with other materials to achieve a variety of functions, which will be similar if their active sites are similar. Thus we can infer a proteinpsilas function by identifying its binding area. This paper proposes a novel method to select a proteinpsilas binding area using the Markov Cluster (MCL) algorithm. A distance matrix is constructed from the surface residues distance on the protein, then transformed to the connectivity matrix for application of the MCL process, and finally evaluated by using Catalytic Site Atlas (CSA) data. In the experimental result using CSA data which comprised 94 selected single chain proteins, our algorithm detects 91 (97%) binding areas near the active site of each protein. We introduced new geometrical features with the aim of improving the prediction accuracy of the active site residues by selecting the residues near the active site.

A compressing method for genome sequence cluster using sequence alignment

After identifying the function of a protein, biologists produce new useful proteins by substituting some residues of the identified protein. These new proteins have high sequence homology (similarity). We define a sequence cluster as a cluster that is constituted of similar sequences. As another example of a sequence cluster, we consider a SNP (single nucleotide polymorphism) cluster. A SNP is a DNA sequence variation occurring when a single nucleotide in the genome (or other shared sequence) differs between members of a species (or between paired chromosomes in an individual). We suggest a new compressing technique for these sequence clusters using a sequence alignment method. We select a representative sequence which has a minimum sequence distance in the cluster by scanning distances of all sequences. The distances are obtained by calculating a sequence alignment score. The result of this sequence alignment is utilized to author conversion information called an edit-script between the two sequences. We only stored representative sequences and edit-scripts of each cluster into a database. Member sequences of each cluster can then be easily created using representative sequences and edit-scripts.