Search

A correction has been published and is appended to both the HTML and PDF versions of this paper. The error has not been fixed in the paper.

A scaffold obtained from whole-genome shotgun sequencing of Paenibacillus popilliae ATCC 14706T shares partial homology with plasmids found in other strains of P. popilliae. PCR and sequencing for gap enclosure indicated that the scaffold originated from a 15,929-bp circular DNA. The restriction patterns of a plasmid isolated from P. popilliae ATCC 14706T were identical to those expected from the sequence; thus, this circular DNA was identified as a plasmid of ATCC 14706T and designated pPOP15.9. The plasmid encodes 17 putative open reading frames. Orfs 1, 5, 7, 8, and 9 are homologous to Orfs 11, 12, 15, 16, and 17, respectively. Orf1 and Orf11 are annotated as replication initiation proteins. Orf8 and Orf16 are homologs of KfrA, a plasmid-stabilizing protein in Gram-negative bacteria. Recombinant Orf8 and Orf16 proteins were assessed for the properties of KfrA. Indeed, they formed multimers and bound to inverted repeat sequences in upstream regions of both orf8 and orf16. A phylogenetic tree based on amino acid sequences of Orf8, Orf16 and Kfr proteins did not correlate with species lineage., Highlights • A 15.9 kb plasmid of P. popilliae was identified and completely sequenced. • The plasmid was predicted to encode 17 putative open reading frames. • Recombinant KfrA proteins formed multimers and bound upstream of the kfrA genes. • Phylogenetic analysis suggests that kfrA genes were horizontally transferred.

To determine the phylogenetic relationship among Paenibacillus species, putative replication origin regions were compared. In the rsmG-gyrA region, gene arrangements in Paenibacillus species were identical to those of Bacillus species, with the exception of an open reading frame (orf14) positioned between gyrB and gyrA, which was observed only in Paenibacillus species. The orf14 product was homologous to the endospore-associated proteins YheC and YheD of Bacillus subtilis. Phylogenetic analysis based on the YheCD proteins suggested that Orf14 could be categorized into the YheC group. In the Paenibacillus genome, DnaA box clusters were found in rpmH-dnaA and dnaA-dnaN intergenic regions, known as box regions C and R, respectively; this localization was similar to that observed in B. halodurans. A phylogenetic tree based on the nucleotide sequences of the whole replication origin regions suggested that P. popilliae, P. thiaminolyticus, and P. dendritiformis are closely related species.

// Koichi Azuma 1 , Akihiko Kawahara 2 , Kahori Sonoda 3 , Kazutaka Nakashima 2 ,Kousuke Tashiro 4 , Kosuke Watari 3 , Hiroto Izumi 5 , Masayoshi Kage 2 , Michihiko Kuwano 6 , Mayumi Ono 3 and Tomoaki Hoshino 1 1 Division of Respirology, Neurology, and Rheumatology, Department of Internal Medicine, Kurume University School of Medicine, Kurume, Fukuoka, Japan 2 Department of Diagnostic Pathology, Kurume University Hospital, Kurume, Fukuoka, Japan 3 Department of Pharmaceutical Oncology, Graduate School of Pharmaceutical Sciences, Kyushu University, Fukuoka, Japan 4 Department of Bioscience and Biotechnology, Kyushu University, Fukuoka, Japan 5 Department of Occupational Pneumology, Institute of Industrial Ecological Sciences, University of Occupational and Environmental Health, Kitakyushu, Japan 6 Laboratory of Molecular Cancer Biology, Department of Clinical Pharmaceutics, Graduate School of Pharmaceutical Sciences, Kyushu University, Fukuoka, Japan Correspondence: Koichi Azuma, email: // Keywords : activating EGFR mutation, FGFR1, non-small cell lung cancer, afatinib Received : January 31, 2014 Accepted : March 24, 2014 Published : March 26, 2014 Abstract Most NSCLC patients with EGFR mutations benefit from treatment with EGFR-TKIs, but the clinical efficacy of EGFR-TKIs is limited by the appearance of drug resistance. Multiple kinase inhibitors of EGFR family proteins such as afatinib have been newly developed to overcome such drug resistance. We established afatinib-resistant cell lines after chronic exposure of activating EGFR mutation-positive PC9 cells to afatinib. Afatinib-resistant cells showed following specific characteristics as compared to PC9: [ 1 ] Expression of EGFR family proteins and their phosphorylated molecules was markedly downregulated by selection of afatinib resistance; [ 2 ] Expression of FGFR1 and its ligand FGF2 was alternatively upregulated; [ 3 ] Treatment with anti-FGF2 neutralizing antibody blocked enhanced phosphorylation of FGFR in resistant clone; [ 4 ] Both resistant clones showed collateral sensitivity to PD173074, a small-molecule FGFR-TKIs, and treatment with either PD173074 or FGFR siRNA exacerbated suppression of both afatinib-resistant Akt and Erk phosphorylation when combined with afatinib; [ 5 ] Expression of twist was markedly augmented in resistant sublines, and twist knockdown specifically suppressed FGFR expression and cell survival. Together, enhanced expression of FGFR1 and FGF2 thus plays as an escape mechanism for cell survival of afatinib-resistant cancer cells, that may compensate the loss of EGFR-driven signaling pathway.

We conducted an exhaustive study of gene expression in fig fruits to identify the gene complexes responsible for fundamental fruit physiology and phenotypic differences between ecotypes. We performed high-throughput pyrosequencing on cDNA libraries constructed from caprifig and common fig fruits and compared their transcriptomes by analyzing the expressed sequence tags obtained. We collected a total of 290,594 expressed sequence tag reads from the two fruit types and assembled them into 71,455 unigenes (19,166 contigs and 52,289 singletons). We identified many metabolic genes, including those encoding proteins in the ethylene, glucose, and anthocyanin synthesis pathways that are involved in fruit maturation. This set also contained unigenes with unidentified functions. We observed no significant differences between the fruit types with respect to Gene Ontology term representation. By reverse transcription polymerase chain reaction, however, we detected several polymorphisms at the level of individual genes. Inter-type variations with respect to the expression level or transcription product size were observed in B- and C-class MADS-box gene homologs and chalcone synthase homologs, which are believed to be involved in sexuality and parthenocarpy, respectively. Expression polymorphisms were also observed for other genes, including a gibberellin-regulated protein gene. Our data and results contribute to genetic research on fig fruits and will aid in the understanding of fruit physiology and mechanisms of phenotypic differentiation.

BACKGROUND: Because only a subset of patients show clinical responses to peptide-based cancer vaccination, it is critical to identify biomarkers for selecting patients who would most likely benefit from this treatment. METHODS: The authors characterized the gene expression profiles in peripheral blood of vaccinated patients to identify biomarkers to predict patient prognosis. Peripheral blood was obtained from advanced castration-resistant prostate cancer patients, who survived for >900 days (long-term survivors, n = 20) or died within 300 days (short-term survivors, n = 20) after treatment with personalized peptide vaccination. Gene expression profiles in prevaccination and postvaccination peripheral blood mononuclear cells (PBMCs) were assessed by DNA microarray. RESULTS: There were no statistically significant differences in the clinical or pathological features between the 2 groups. Microarray analysis of prevaccination PBMCs identified 19 genes that were differentially expressed between the short-term and long-term survivors. Among the 15 up-regulated genes in the short-term survivors, 13 genes, which were also differentially expressed in postvaccination PBMCs, were associated with gene signatures of granulocytes. When a set of 4 differentially expressed genes were selected as the best combination to determine patient survival, prognosis was correctly predicted in 12 of 13 patients in a validation set (accuracy, 92%). CONCLUSIONS: These results suggested that abnormal granulocytes present in the PBMC faction may contribute to poor prognosis in advanced prostate cancer patients receiving personalized peptide vaccination. Gene expression profiling in peripheral blood might thus be informative for devising better therapeutic strategies by predicting patient prognosis after cancer vaccines. Cancer 2012;118: 3208–21. © 2011 American Cancer Society.

Clostridium perfringens, a Gram-positive anaerobic pathogen, is a causative agent of human gas gangrene that leads to severe rapid tissue destruction and can cause death within hours unless treated immediately. Production of several toxins is known to be controlled by the two-component VirR/VirS system involving a regulatory RNA (VR-RNA) in C. perfringens. To elucidate the precise regulatory network governed by VirR/VirS and VR-RNA, a series of microarray screening using VirR/VirS and VR-RNA-deficient mutants was performed. Finally, by qRT-PCR analysis, 147 genes (30 single genes and 21 putative operons) were confirmed to be under the control of the VirR/VirS-VR-RNA regulatory cascade. Several virulence-related genes for alpha-toxin, kappa-toxin, hyaluronidases, sialidase, and capsular polysaccharide synthesis were found. Furthermore, some genes for catalytic enzymes, various genes for transporters, and many genes for energy metabolism were also found to be controlled by the cascade. Our data indicate that the VirR/VirS-VR-RNA system is a global gene regulator that might control multiple cellular functions to survive and multiply in the host, which would turn out to be a lethal flesh-eating infection.

Previous cell cycle studies have been based on cell-nuclear proliferation only. Eukaryotic cells, however, have double membranes-bound organelles, such as the cell nucleus, mitochondrion, plastids and single-membrane-bound organelles such as ER, the Golgi body, vacuoles (lysosomes) and microbodies. Organelle proliferations, which are very important for cell functions, are poorly understood. To clarify this, we performed a microarray analysis during the cell cycle of Cyanidioschyzon merolae. C. merolae cells contain a minimum set of organelles that divide synchronously. The nuclear, mitochondrial and plastid genomes were completely sequenced. The results showed that, of 158 genes induced during the S or G2-M phase, 93 were known and contained genes related to mitochondrial division, ftsZ1-1, ftsz1-2 and mda1, and plastid division, ftsZ2-1, ftsZ2-2 and cmdnm2. Moreover, three genes, involved in vesicle trafficking between the single-membrane organelles such as vps29 and the Rab family protein, were identified and might be related to partitioning of single-membrane-bound organelles. In other genes, 46 were hypothetical and 19 were hypothetical conserved. The possibility of finding novel organelle division genes from hypothetical and hypothetical conserved genes in the S and G2-M expression groups is discussed.

Asia differs substantially among and within its regions populated by diverse ethnic groups, which maintain their own respective cultures and dietary habits. To address the diversity in their gut microbiota, we characterized the bacterial community in fecal samples obtained from 303 school-age children living in urban or rural regions in five countries spanning temperate and tropical areas of Asia. The microbiota profiled for the 303 subjects were classified into two enterotype-like clusters, each driven by Prevotella (P-type) or Bifidobacterium/Bacteroides (BB-type), respectively. Majority in China, Japan and Taiwan harbored BB-type, whereas those from Indonesia and Khon Kaen in Thailand mainly harbored P-type. The P-type microbiota was characterized by a more conserved bacterial community sharing a greater number of type-specific phylotypes. Predictive metagenomics suggests higher and lower activity of carbohydrate digestion and bile acid biosynthesis, respectively, in P-type subjects, reflecting their high intake of diets rich in resistant starch. Random-forest analysis classified their fecal species community as mirroring location of resident country, suggesting eco-geographical factors shaping gut microbiota. In particular, children living in Japan harbored a less diversified microbiota with high abundance of Bifidobacterium and less number of potentially pathogenic bacteria, which may reflect their living environment and unique diet.

We prepared microarrays that contain genomic sequences of a heterocyst-forming filamentous cyanobacterium Anabaena sp. PCC 7120. The complete genome of this cyanobacterium codes for about 5,368 protein-coding genes in the main chromosome of 6.4 Mbp. In total, 2,407 DNA segments were selected from the sequencing clones, and amplified by PCR, then spotted on glass slides in duplicate. These microarrays differ from the widely used commercial or custom-made ones for other microorganisms in that each DNA segment was 3-4 kbp long, and contained about 3-4 predicted genes on average. This feature, however, did not decrease the usefulness of the microarrays, since we were able to detect a number of potentially novel genes that are induced in response to nitrogen deprivation, low temperature and drought. In addition, we found some genomic regions in which dozens of contiguous genes are simultaneously regulated. These results suggest that these segment-based microarrays are useful especially for such large genomes as Anabaena, for which the number of genes exceeds either technical or practical limitations.

We propose a new method for identifying and validating drug targets by using gene networks, which are estimated from cDNA microarray gene expression profile data. We created novel gene disruption and drug response microarray gene expression profile data libraries for the purpose of drug target elucidation. We use two types of microarray gene expression profile data for estimating gene networks and then identifying drug targets. The estimated gene networks play an essential role in understanding drug response data and this information is unattainable from clustering methods, which are the standard for gene expression analysis. In the construction of gene networks, we use the Bayesian network model. We use an actual example from analysis of the Saccharomyces cerevisiae gene expression profile data to express a concrete strategy for the application of gene network information to drug discovery.

Gene regulatory networks developed from full genome expression libraries from gene perturbation variant cell lines can be used to quickly and efficiently identify the molecular mechanism of action of drugs or lead compound molecules. We developed an extensive yeast gene expression library consisting of full-genome cDNA array data for over 500 yeast strains each with a single gene disruption. Using this data, combined with dose and time course expression experiments with the oral antifungal agent, we used Boolean network discovery techniques to determine the genes whose expression was most profoundly affected by this drug. Our system identified the gene as the most significantly suppressed target molecule due to exposure to the antifungal agent. This process for network based drug discovery can significantly decrease the time and resources necessary to make rational drug targeting decisions.

In the present study, we determined the draft genome sequence of the entomopathogenic bacterium Serratia liquefaciens FK01, which is highly virulent to the silkworm. The draft genome is ~5.28 Mb in size, and the G+C content is 55.8%.

Because only a subset of patients show clinical responses to peptide-based cancer vaccination, it is critical to identify biomarkers for selecting patients who would most likely benefit from this treatment.The authors characterized the gene expression profiles in peripheral blood of vaccinated patients to identify biomarkers to predict patient prognosis. Peripheral blood was obtained from advanced castration-resistant prostate cancer patients, who survived for900 days (long-term survivors, n = 20) or died within 300 days (short-term survivors, n = 20) after treatment with personalized peptide vaccination. Gene expression profiles in prevaccination and postvaccination peripheral blood mononuclear cells (PBMCs) were assessed by DNA microarray.There were no statistically significant differences in the clinical or pathological features between the 2 groups. Microarray analysis of prevaccination PBMCs identified 19 genes that were differentially expressed between the short-term and long-term survivors. Among the 15 up-regulated genes in the short-term survivors, 13 genes, which were also differentially expressed in postvaccination PBMCs, were associated with gene signatures of granulocytes. When a set of 4 differentially expressed genes were selected as the best combination to determine patient survival, prognosis was correctly predicted in 12 of 13 patients in a validation set (accuracy, 92%).These results suggested that abnormal granulocytes present in the PBMC faction may contribute to poor prognosis in advanced prostate cancer patients receiving personalized peptide vaccination. Gene expression profiling in peripheral blood might thus be informative for devising better therapeutic strategies by predicting patient prognosis after cancer vaccines.

Mechanisms of recovery from heat injury in Salmonella typhimurium were elucidated. Recovery of the heat-injured S. typhimurium cells in TSB resulted in full recovery after 3 h of incubation at 37°C. The DNA microarray analysis of 30- and 60-min recovering cells resulted in an increase in transcription of 89 and 141 genes, respectively. Among them, 15 genes, with known function, seemed to be somewhat involved in recovery. They encoded proteins involved in branched-chain amino acid (BCAA) transport (livJ, livH), cell envelope integrity (ddg), heat-shock response (cpxP, rrmJ), phage shock protein (pspA), ribosome modulation factor (rmf), virulence (sseB) transcriptional regulation (rpoE, rpoH, rseA, rseB, rseC) and ArcB signal transduction (sixA) and cytoplasmic membrane protein (fxsA). Among them, the effects of BCAA supplementation on recovery from heat injury were studied to confirm the importance of the BCAA transport liv genes during recovery. It was found that supplementation of TSB with 0.1% BCAA resulted in an enhanced recovery of injured cells in comparison to those recovered in TSB without BCAA. Supplementation of BCAA at 0.1% resulted in a cell count increase 4.4-fold greater than that of the control after 1 h incubation. It seems that BCAA promoted the recovery by promoting protein synthesis either directly through their use in translation or indirectly through stimulation of protein synthesis by activation of the Lrp protein.

Some drugs affect secretion of secreted proteins (e.g. cytokines) released from target cells, but it remains unclear whether these proteins act in an autocrine manner and directly effect the cells on which the drugs act. In this study, we propose a computational method for testing a biological hypothesis: there exist autocrine signaling pathways that are dynamically regulated by drug response transcriptome networks and control them simultaneously. If such pathways are identified, they could be useful for revealing drug mode-of-action and identifying novel drug targets. By the node-set separation method proposed, dynamic structural changes can be embedded in transcriptome networks that enable us to find master-regulator genes or critical paths at each observed time. We then combine the protein-protein interaction network with the estimated dynamic transcriptome network to discover drug-affected autocrine pathways if they exist. The statistical significance (p-values) of the pathways are evaluated by the meta-analysis technique. The dynamics of the interactions between the transcriptome networks and the signaling pathways will be shown in this framework. We illustrate our strategy by an application using anti-hyperlipidemia drug, Fenofibrate. From over one million protein-protein interaction pathways, we extracted significant 23 autocrine-like pathways with the Bonferroni correction, including VEGF-NRP1-GIPC1-PRKCA-PPARalpha, that is one of the most significant ones and contains PPARalpha, a target of Fenofibrate.

Fenofibrate is a synthetic ligand for the nuclear receptor peroxisome proliferator-activated receptor (PPAR) alpha and has been widely used in the treatment of metabolic disorders, especially hyperlipemia, due to its lipid-lowering effect. The molecular mechanism of lipid-lowering is relatively well defined: an activated PPARalpha forms a PPAR-RXR heterodimer and this regulates the transcription of genes involved in energy metabolism by binding to PPAR response elements in their promoter regions, so-called "trans-activation". In addition, fenofibrate also has anti-inflammatory and anti-athrogenic effects in vascular endothelial and smooth muscle cells. We have limited information about the anti-inflammatory mechanism of fenofibrate; however, "trans-repression" which suppresses production of inflammatory cytokines and adhesion molecules probably contributes to this mechanism. Furthermore, there are reports that fenofibrate affects endothelial cells in a PPARalpha-independent manner. In order to identify PPARalpha-dependently and PPARalpha-independently regulated transcripts, we generated microarray data from human endothelial cells treated with fenofibrate, and with and without siRNA-mediated knock-down of PPARalpha. We also constructed dynamic Bayesian transcriptome networks to reveal PPARalpha-dependent and -independent pathways. Our transcriptome network analysis identified growth differentiation factor 15 (GDF15) as a hub gene having PPARalpha-independently regulated transcripts as its direct downstream children. This result suggests that GDF15 may be PPARalpha-independent master-regulator of fenofibrate action in human endothelial cells.

Monoclonal cell populations are known to be composed of heterogeneous subpopulations, thus complicating the data analysis. To gain clear insights into the mechanisms of cellular systems, biological data from a homogeneous cell population should be obtained. In this study, we developed a method based on Latent Profile Analysis (LPA) combined with Confirmatory Factor Analysis (CFA) to divide mixed data into classes, depending on their heterogeneity. In general cluster analysis, the number of measured points is a constraint, and thereby the data must be classified into fewer groups than the number of samples. By our newly developed method, the measured data can be divided into groups depending on their latent effects, without constraints. Our method is useful to clarify all types of omics data, including transcriptome, proteome and metabolic information.

Escherichia coli O157, an etiological agent of hemorrhagic colitis and hemolytic uremic syndrome, is one of the leading worldwide public health threats. Genome sequencing of two O157 strains have revealed that the chromosome is comprised of a 4.1 Mb backbone shared by K-12 and a total of 1.4 Mb O157-specific sequences. Most of the large O157-specific sequences are prophages and prophage-like elements, which have carried many virulence genes into the O157 genome. This suggests that bacteriophages have played the key roles in the emergence of O157. The Whole Genome PCR Scanning (WGPScanning) analysis of O157 strains, on the other hand, revealed a high level of genomic diversity in O157. Variation of prophages has also been suggested as a major factor generating such diversity. In this study, we analyzed the gene content of O157 strains, by an oligoDNA microarray, using the same set of strains as examined by the WGPScanning method. Although most of the strains were typical O157 : H7, they differed remarkably in gene composition, particularly in those on prophages, and we identified more than 400 'variably absent or present' genes which included virulence-related genes. This confirms the role of prophages in generating the genomic diversity, and raises a possibility that some level of variation in potential virulence is present among O157 strains. Fine comparison of the two datasets obtained by microarray and WGPScanning provided much further details on the O157 genome diversity than illustrated by each method alone, indicating the usefulness of this combinational approach in the genomic comparison of closely related strains.

We present a computational method for identifying genes and their regulatory pathways influenced by a drug, using microarray gene expression data collected by single gene disruptions and drug responses. The automatic identification of such genes and pathways in organisms' cells is an important problem for pharmacogenomics and the tailor-made medication. Our method estimates regulatory relationships between genes as a gene network from microarray data of gene disruptions with a Bayesian network model, then identifies the drug affected genes and their regulatory pathways on the estimated network with time course drug response microarray data. Compared to the existing method, our proposed method can identify not only the drug affected genes and the druggable genes, but also the drug responses of the pathways. For evaluating the proposed method, we conducted simulated examples based on artificial networks and expression data. Our method succeeded in identifying the pseudo drug affected genes and pathways with the high coverage greater than 80 %. We also applied our method to Saccharomyces cerevisiae drug response microarray data. In this real example, we identified the genes and the pathways that are potentially influenced by a drug. These computational experiments indicate that our method successfully identifies the drug-activated genes and pathways, and is capable of predicting undesirable side effects of the drug, identifying novel drug target genes, and understanding the unknown mechanisms of the drug.

We propose a computational strategy for discovering gene networks affected by a chemical compound. Two kinds of DNA microarray data are assumed to be used: One dataset is short time-course data that measure responses of genes following an experimental treatment. The other dataset is obtained by several hundred single gene knock-downs. These two datasets provide three kinds of information; (i) A gene network is estimated from time-course data by the dynamic Bayesian network model, (ii) Relationships between the knocked-down genes and their regulatees are estimated directly from knock-down microarrays and (iii) A gene network can be estimated by gene knock-down data alone using the Bayesian network model. We propose a method that combines these three kinds of information to provide an accurate gene network that most strongly relates to the mode-of-action of the chemical compound in cells. This information plays an essential role in pharmacogenomics. We illustrate this method with an actual example where human endothelial cell gene networks were generated from a novel time course of gene expression following treatment with the drug fenofibrate, and from 270 novel gene knock-downs. Finally, we succeeded in inferring the gene network related to PPAR-alpha, which is a known target of fenofibrate.

The Wnt signaling pathway is activated in most human colorectal tumors. Mutational inactivation in the tumor suppressor adenomatous polyposis coli (APC), as well as activation of beta-catenin, causes the accumulation of beta-catenin, which in turn associates with the T cell factor/lymphoid enhancer factor (TCF/LEF) family of transcription factors and activates transcription of their target genes. Here we show that beta-catenin activates transcription of the BMP and activin membrane-bound inhibitor (BAMBI)/NMA gene. The expression level of BAMBI was found to be aberrantly elevated in most colorectal and hepatocellular carcinomas relative to the corresponding non-cancerous tissues. Expression of BAMBI in colorectal tumor cell lines was repressed by a dominant-negative mutant of TCF-4 or by an inhibitor of beta-catenin-TCF interaction, suggesting that beta-catenin is responsible for the aberrant expression of BAMBI in colorectal tumor cells. Furthermore, overexpression of BAMBI inhibited the response of tumor cells to transforming growth factor-beta signaling. These results suggest that beta-catenin interferes with transforming growth factor-beta-mediated growth arrest by inducing the expression of BAMBI, and this may contribute to colorectal and hepatocellular tumorigenesis.

Gene regulatory networks developed from full genome expression libraries from gene perturbation variant cell lines can be used to quickly and efficiently identify the molecular mechanism of action of drugs or lead compound molecules. We developed an extensive yeast gene expression library consisting of full-genome cDNA array data for over 500 yeast strains each with a single gene disruption. Using this data, combined with dose and time course expression experiments with the oral antifungal agent, we used Boolean network discovery techniques to determine the genes whose expression was most profoundly affected by this drug. Our system identified the gene as the most significantly suppressed target molecule due to exposure to the antifungal agent. This process for network based drug discovery can significantly decrease the time and resources necessary to make rational drug targeting decisions.

We propose a new statistical method for constructing a genetic network from microarray gene expression data by using a Bayesian network. An essential point of Bayesian network construction is the estimation of the conditional distribution of each random variable. We consider fitting nonparametric regression models with heterogeneous error variances to the microarray gene expression data to capture the nonlinear structures between genes. Selecting the optimal graph, which gives the best representation of the system among genes, is still a problem to be solved. We theoretically derive a new graph selection criterion from Bayes approach in general situations. The proposed method includes previous methods based on Bayesian networks. We demonstrate the effectiveness of the proposed method through the analysis of Saccharomyces cerevisiae gene expression data newly obtained by disrupting 100 genes.

Comparing the genomes of O157 and non-O157 enterohemorrhagic Escherichia coli (EHEC) strains reveals the selective conservation of a large number of virulence determinants., Background Enterohemorrhagic Escherichia coli (EHEC) O157 causes severe food-borne illness in humans. The chromosome of O157 consists of 4.1 Mb backbone sequences shared by benign E. coli K-12, and 1.4 Mb O157-specific sequences encoding many virulence determinants, such as Shiga toxin genes (stx genes) and the locus of enterocyte effacement (LEE). Non-O157 EHECs belonging to distinct clonal lineages from O157 also cause similar illness in humans. According to the 'parallel' evolution model, they have independently acquired the major virulence determinants, the stx genes and LEE. However, the genomic differences between O157 and non-O157 EHECs have not yet been systematically analyzed. Results Using microarray and whole genome PCR scanning analyses, we performed a whole genome comparison of 20 EHEC strains of O26, O111, and O103 serotypes with O157. In non-O157 EHEC strains, although genome sizes were similar with or rather larger than O157 and the backbone regions were well conserved, O157-specific regions were very poorly conserved. Around only 20% of the O157-specific genes were fully conserved in each non-O157 serotype. However, the non-O157 EHECs contained a significant number of virulence genes that are found on prophages and plasmids in O157, and also multiple prophages similar to, but significantly divergent from, those in O157. Conclusion Although O157 and non-O157 EHECs have independently acquired a huge amount of serotype- or strain-specific genes by lateral gene transfer, they share an unexpectedly large number of virulence genes. Independent infections of similar but distinct bacteriophages carrying these virulence determinants are deeply involved in the evolution of O157 and non-O157 EHECs.