Current genome projects are resulting in a flood of sequence data. The interpretation of these sequences is lagging, and optimized data analysis strategies need to be developed. Much can be learned from comparing different genomes, as genomes of distant organisms may still encode proteins with high sequence similarity. The order of genes (co linearity) in genomes may also be conserved to some extend. We have employed both these observations to create a multi-functional, computational analysis system (genomeSCOUT) which allows for rapid identification and functional characterization of genes and proteins through genome comparison. With a number of independent algorithms, information about different levels of protein homology (concerning e.g. paralogs, orthologs and clusters of orthologous groups, COGs) and gene order is collected and stored in several value added databases. These databases are then used for interactive comparison of genomes and subsequent analysis. The application is based on the well established data integration system SRS. This ensures (1) fast handling of large genomic data sets, (2) straightforward access to a multitude of biological databases, (3) unique linking functions between these databases, (4) highly efficient collection of information on genes and proteins, and 5. fully integrated and user friendly graphical representations of search results. This application can be used for projects as diverse as the correct annotation of genomes, the optimization of (micro) organisms for industrial production, or the identification of drug targets.
In this paper, we propose and use two novel procedures for the analysis of microarray gene expression data. The first is correspondence analysis which visualizes the relationship between genes and tissues as two 2 dimensional graphs, oriented so that distances between genes are preserved, distances between tissues are preserved, and so that genes which primarily distinguish certain types of tissue are spatially close to those tissues. For the inference of genetic links, partial correlations rather than correlations are the key issue. A partial correlation between i and j is the relationship between i and j after the effect of surrounding genes has been subtracted out of their pairwise correlation. This leads to the area of graphical modeling. A limitation of the graphical modeling approach is that the correlation matrix of expression profiles between genes is degenerate whenever the number of genes to be analyzed exceeds the number of distinct expression measurements. This can cause considerable problems, as calculation of partial correlations typically uses the inverse of the correlation matrix. To avoid this limitation, we propose two practical multiple regression procedures with variable selection to measure the net, screened, relationship between pairs of genes. Possible biases arising from the analysis of a subset of genes from the genome are examined in the worked examples. It seems that both these approaches are more natural ways of analyzing gene expression data than the currently popular approach of two way clustering.
We have implemented a system for assisting experts in selecting MEDLINE records for database construction purposes. This system has two specific features: The first is a learning mechanism which extracts characteristics in the abstracts of MEDLINE records of interest as patterns. These patterns reflect selection decisions by experts and are used for screening the records. The second is a keyword recommendation system which assists and supplements experts' knowledge in unexpected cases. Combined with a conventional keyword-based information retrieval system, this system may provide an efficient and comfortable environment for MEDLINE record selection by experts. Some computational experiments are provided to prove that this idea is useful.
Today, more and more DNA sequences are becoming available. The information about DNA sequences are stored in molecular biology databases. The size and importance of these databases will be bigger and bigger in the future, therefore this information must be stored or communicated efficiently. Furthermore, sequence compression can be used to define similarities between biological sequences. The standard compression algorithms such as gzip or compress cannot compress DNA sequences, but only expand them in size. On the other hand, CTW (Context Tree Weighting Method) can compress DNA sequences less than two bits per symbol. These algorithms do not use special structures of biological sequences. Two characteristic structures of DNA sequences are known. One is called palindromes or reverse complements and the other structure is approximate repeats. Several specific algorithms for DNA sequences that use these structures can compress them less than two bits per symbol. In this paper, we improve the CTW so that characteristic structures of DNA sequences are available. Before encoding the next symbol, the algorithm searches an approximate repeat and palindrome using hash and dynamic programming. If there is a palindrome or an approximate repeat with enough length then our algorithm represents it with length and distance. By using this preprocessing, a new program achieves a little higher compression ratio than that of existing DNA-oriented compression algorithms. We also describe new compression algorithm for protein sequences.
To accelerate the calculation speed for simulating a biological system, we proposed a novel simulation method, the two-phase partition method, which calculated molecular processes at a higher speed than any other proposed method. This method divides a biological system, which can be described by chemical reaction equations, into two-phases: the binding and reaction phases. We demonstrated the capability of the two-phase partition method to simulate a complex biological system at an extremely high speed and clarified the accuracy of the simulation. The two-phase partition method is very useful for simulating complex interactions among proteins and DNAs.
Annotation and database system of full-length cDNA sequences was developed. As the components of the system, ORF annotation system, functional annotation system based on database search results, mapping annotation system, and integrated retrieval and display system were developed. In the ORF annotation system integrated analyses using conventional tools are performed and useful retrieval interface using motif list are introduced. In the functional annotation system based on database search results, a new method that characterizes a given unknown cDNA was developed by using a profile of similarity level over words appearing in sequence database entries. In the mapping annotation system, we linked by similarity searches full-length cDNA sequences with database DNA sequences that are already mapped on chromosomes. By using these links, full-length cDNAs can be retrieved by the retrieval condition of physical mapping information. Genetic disease information mapped on the physical mapping site can also be displayed by this system. Furthermore, we constructed an integrated database system for these analyzed data, and thus enabled annotation and selection of full-length cDNAs from points of both gene function and mapping information.
Inferring gene regulatory networks by differential equations from the time series data of a DNA microarray is one of the most challenging tasks in the post-genomic era. However, there have been no studies actually inferring gene regulatory networks by differential equations from genome-level data. The reason for this is that the number of parameters in the equations exceeds the number of measured time points. We here succeeded in executing the inference, not by directly determining parameters but by applying multiple regression analysis to our equations. We derived our differential equations and steady state equations from the rate equations of transcriptional reactions in an organism. Verification with a number of genes related to respiration indicated the validity and effectiveness of our method. Moreover, the steady state equations were more appropriate than the differential equations for the microarray data used.
Intrinsic protein disorder refers to segments or to whole proteins that fail to fold completely on their own. Here we predicted disorder on protein sequences from 34 genomes, including 22 bacteria, 7 archaea, and 5 eucaryotes. Predicted disordered segments > or = 50, > or = 40, and > or = 30 in length were determined as well as proteins estimated to be wholly disordered. The five eucaryotes were separated from bacteria and archaea by having the highest percentages of sequences predicted to have disordered segments > or = 50 in length: from 25% for Plasmodium to 41% for Drosophila. Estimates of wholly disordered proteins in the bacteria ranged from 1% to 8%, averaging to 3 +/- 2%, estimates in various archaea ranged from 2 to 11%, plus an apparently anomalous 18%, averaging to 7 +/- 5% that drops to 5 +/- 3% if the high value is discarded. Estimates in the 5 eucarya ranged from 3 to 17%. The putative wholly disordered proteins were often ribosomal proteins, but in addition about equal numbers were of known and unknown function. Overall, intrinsic disorder appears to be a common, with eucaryotes perhaps having a higher percentage of native disorder than archaea or bacteria.
The most challenging strategy for analyzing genome-wide polymorphisms and/or expression profiles is to solve multi-factor causal-relationship simultaneously. As the first step, we propose a framework of association study using maximum likelihood method that simultaneously handles genetic polymorphisms and epi-genetic information, e.g. environmental factors. We evaluate the theory by applying it to genotyped data of myocardial infarction (MI) patients.
Single particle analysis is a straightforward method for studying the structures of macromolecules that cannot be crystallized. It builds three-dimensional structures of particles by estimating the projection angles of their randomly oriented electron-microscopic images. The existing methods divide the images into clusters, build class averages for the clusters, and estimate the projection angle of each cluster. However, the clustering and the averaged images are highly sensitive to the choice of reference images and mask patterns for each cluster. Thus, the analyses are neither robust nor automatic, and their results depend heavily on the intuition and experience of researchers who set references. We have been developing a software system for single-particle analysis with new clustering and averaging algorithms for building the three-dimensional structures of target molecules. In this paper, we focus on the algorithms for the robust image-processing of the electron microscopic images in our system.
More than 6,000 amino acid sequence attributes were ranked by their conditional probabilities for indicating ordered or disordered protein structure. The top 10 each from several different groups of attributes were merged with still other attributes and then subjected to selection by logistic regression. Evidently, the determination of order or disorder results from the interplay among several attributes, such as average Coordination Number, aromatic content and the numbers of non-polar amino acids, all of which favor the ordered state, and others like Net Charge, Flexibility Index, and the presence of certain polar amino acids, all of which favor disorder. The top 12 selected attributes were used as inputs for artificial neural network (ANN) predictors. Five predictors were developed, compared with each other, and with previous work. The best of these shows substantially improved generalization compared to our previously published predictor.
A protein sequence-structure alignment method for database searches is examined on how effectively this method together with a simple scoring function previously developed can identify compatibilities between sequences and structures of proteins. The scoring function consists of pairwise contact energies, repulsive packing potentials of residues for overly dense arrangement and short-range potentials for secondary structures. Pairwise contact interactions in a sequence-structure alignment are evaluated in a mean field approximation on the basis of probabilities of site pairs to be aligned. Gap penalties are assumed to be proportional to the number of contacts at each residue position, and as a result gaps will be more frequently placed on protein surfaces than in cores. In addition to minimum energy alignments, we use probability alignments made by successively aligning site pairs in order by pairwise alignment probabilities. Results show that the present energy function and alignment method can detect well both folds compatible with a given sequence and, inversely, sequences compatible with a given fold. Probability alignments consisting of most reliable site pairs only can yield small root mean square deviations, and including less reliable pairs increases the deviations. Remarkably, by this method some individual sequence-structure pairs are detected having only 5-20% sequence identity.
We propose a new type of DNA chips with logical operations executable. In DNA computing researches, some methods have been developed to represent and evaluate Boolean functions on DNA strands. By employing the evaluation methods, we are able to deal with logical operations such as logical-and and logical-or for gene expressions. By combining with the DNA Coded Number method, we implement universal DNA chips which not only detect gene expressions but also find logical formulae of gene expressions. An important advantage of our intelligent DNA chip is that the intensity of the fluorescence at each element is not only proportional to the expression level of the genes in the sample but also proportional to the satisfiability level of the Boolean formula at the element with the gene expression pattern. These features of the intelligent DNA chip and the DCN method allow us more quantitative analyses of gene expression profiles and the logical operations.
DNA sequencing is a very important problem in genomics. Several different sequencing methods are currently utilized. One promising method uses a sequencing chip to obtain information about the presence of subsequences in DNA. This paper deals with sequencing of hybridization data from a sequencing chip, called Sequencing by Hybridization (SBH). Preparata et al. proposed a new sequencing chip using universal bases, together with a new sequencing algorithm, and showed that its performance is significantly better than the standard scheme based on oligomer probes. However, the presence of errors in the sequencing chip was not considered, and the method of Preparata et al. cannot be used directly in practice. This paper proposes sequencing algorithms in the presence of hybridization errors for their sequencing chip and applies these algorithms to random data in the presence of random errors. Computational results show that false negative errors have larger effects on the rates of correct reconstruction than do false positive errors. Our extended sequencing algorithms are useful when there are a few hybridization errors.
We propose a parallel hybrid genetic algorithm for flexible protein-protein docking in order to improve the conventional rigid-body models to manipulate protein-protein interactions. The proposed hybrid algorithm is a combination of an evolutionary algorithm with a simulated annealing one, yielding a powerful protein-complex conformation-searching engine. Parallelization of the procedure makes possible to reach high algorithm performance, in both, execution times and size of treated monomers and complexes. Knowledge on side chain flexibility is extracted by means of an exhaustive analysis of crystallographic data on proteins and protein complexes. Results demonstrate the competency of the algorithm since comparison of calculated and crystallographic data accounts for a maximum of 2.5A in RMS difference, including side chain conformation. The system allows routine analysis of this fundamental molecular biology problem important to elucidate bio-macromolecular function in biophysical and biochemical mechanisms involving molecular recognition and interaction, yielding simultaneously clues for designing new proteins and enzymes directed to different purposes.
The Pedigree Visualizer is a system for visualization of pedigree diagrams. It accepts a simple text-based specification of a pedigree diagram. The pedigree diagram is then layout automatically. Both GIF- and PS-formatted output files are produced. In addition, the Pedigree Visualizer also provides a rich set of functions for the manipulation and management of large pedigree files.
As rice genomicsdata continue to accumulate at a rapid rate, are becoming more valuable to warehouse and access large and rigorous data sets. This article gives an overview of available resources on rice bioinformatics and their role in elucidating and propagating biological and genomic information in rice. Of particular focus here is the informatics infrastructure developed at the Rice Genome Research Program (RGP) following an extensive rice genome analysis. The database named INE (Integrated Rice Genome Explorer) integrates the genetic and physical mapping information with the genome sequence being generated in collaboration with the International Rice Genome Sequencing Project (IRGSP). Database links are initially evaluated using an interoperable query tool to explore and compare data across the rice and maize genome databases and potential application to multiple crop database querying. A proposed logistics for interlinking these resources is presented to integrate, manipulate and analyze information on the rice genome. One of the biggest challenges of rice bioinformatics lies in the emerging role of rice as a model system among grass crop species. In view of the importance of comparative genetics in the formulation of new knowledge on plant genomes and genes, comparative bioinformatics remains an essential strategy to gain new insights on the needs and expectations on rice genomics.
Expression arrays facilitate the monitoring of changes in expression patterns of large collections of genes. It is generally expected that genes with similar expression patterns would correspond to proteins of common biological function. We assess this common assumption by comparing levels of similarity of expression patterns and statistical significance of biological terms that describe the corresponding protein functions. Terms are automatically obtained by mining large collections of Medline abstracts. We propose that the combined use of the tools for expression profiles clustering and automatic function retrieval, can be useful tools for the detection of biologically relevant associations between genes in complex gene expression experiments. The results obtained using publicly available experimental data show how, in general, an increase in the similarity of the expression patterns is accompanied by an enhancement of the amount of specific functional information or, in other words, how the selected terms became more specific following an increase in the specificity of the expression patterns. Particularly interesting are the discrepancies from this general trend, i.e. groups of genes with similar expression patterns but very little in common at the functional level. In these cases the similarity of their expression profiles becomes the first link between previously unrelated genes.
In this paper we introduce a new inference method of a gene regulatory network from steady-state gene expression data. Our method determines a regulatory structure consistent with an observed set of steady-state expression profiles, each generated from wild-type and single deletion mutant of the target network. Our method derives the regulatory relationships in the network using a graph theoretic approach. The advantage of our method is to be able to deal with continuous values of steady-state data, while most of the methods proposed in past use a Boolean network model with binary data. Performance of our method is evaluated on simulated networks with varying the size of networks, indegree of each gene, and the data characteristics (continuous-value/binary), and is compared with that of predictor method proposed by Ideker et al. As a result, we show the superiority of using continuous values to binary values, and the performance of our method is much better than that of the predictor method.
At present, there is a lack of a sound methodology to infer causal gene expression relationships on a genome wide basis. We address this first by examining the behaviour of some of the latest and fastest algorithms for tree and cluster analysis, particularly hierarchical methods popular in phylogenetics. Combined with these are two novel distances based on partial, rather than full, correlations. Theoretically, partial correlations should provide better evidence for regulatory genetic links than standard correlations. To compare the clusters obtained by many alternative methods we use tree consensus methods. To compare methods of analysis we used tree partition metrics followed by another level of clustering. These, and a tree fit metric, all suggest that the new distances give quite different trees than those usually obtained. In the second part we consider graphical modeling of the interactions of important genes of the cell cycle. Despite the models seeming to fit well on occasions, and despite the experimental error structure seeming close to multivariate normal, there are considerable problems to overcome. Latent variables, in this case important genes missing from the analysis, are inferred to have a strong effect on the partial correlations. Also, the data show clear evidence of sampling distributions conditional on the status of important cancer related genes, including TP53. Without full information on which genes are wild type the appropriate models cannot be fitted. These findings point to the need to include and distinguish not only all relevant genes but also all splice variants in the design phase of a microarray analysis. Failure to do so will induce problems similar to both latent variables and conditional distributions.
