Knowledge-based analysis of microarray gene expression data by using support vector machines

We introduce a method of functionally classifying genes by using gene expression data from DNA microarray hybridization experiments. The method is based on the theory of support vector machines (SVMs). SVMs are considered a supervised computer learning method because they exploit prior knowledge of gene function to identify unknown genes of similar function from expression data. SVMs avoid several problems associated with unsupervised clustering methods, such as hierarchical clustering and self-organizing maps. SVMs have many mathematical features that make them attractive for gene expression analysis, including their flexibility in choosing a similarity function, sparseness of solution when dealing with large data sets, the ability to handle large feature spaces, and the ability to identify outliers. We test several SVMs that use different similarity metrics, as well as some other supervised learning methods, and find that the SVMs best identify sets of genes with a common function using expression data. Finally, we use SVMs to predict functional roles for uncharacterized yeast ORFs based on their expression data.

Generation of longer cDNA fragments from serial analysis of gene expression tags for gene identification

We have developed a technique called the generation of longer cDNA fragments from serial analysis of gene expression (SAGE) tags for gene identification (GLGI), to convert SAGE tags of 10 bases into their corresponding 3′ cDNA fragments covering hundred bases. A primer containing the 10-base SAGE tag is used as the sense primer, and a single base anchored oligo(dT) primer is used as an antisense primer in PCR, together with Pfu DNA polymerase. By using this approach, a cDNA fragment extending from the SAGE tag toward the 3′ end of the corresponding sequence can be generated. Application of the GLGI technique can solve two critical issues in applying the SAGE technique: one is that a longer fragment corresponding to a SAGE tag, which has no match in databases, can be generated for further studies; the other is that the specific fragment corresponding to a SAGE tag can be identified from multiple sequences that match the same SAGE tag. The development of the GLGI method provides several potential applications. First, it provides a strategy for even wider application of the SAGE technique for quantitative analysis of global gene expression. Second, a combined application of SAGE/GLGI can be used to complete the catalogue of the expressed genes in human and in other eukaryotic species. Third, it can be used to identify the 3′ cDNA sequence from any exon within a gene. It can also be used to confirm the reality of exons predicted by bioinformatic tools in genomic sequences. Fourth, a combined application of SAGE/GLGI can be applied to define the 3′ boundary of expressed genes in the genomic sequences in human and in other eukaryotic genomes.

Human PEX19: cDNA cloning by functional complementation, mutation analysis in a patient with Zellweger syndrome, and potential role in peroxisomal membrane assembly

At least 11 complementation groups (CGs) have been identified for the peroxisome biogenesis disorders (PBDs) such as Zellweger syndrome, for which seven pathogenic genes have been elucidated. We have isolated a human PEX19 cDNA (HsPEX19) by functional complementation of peroxisome deficiency of a mutant Chinese hamster ovary cell line, ZP119, defective in import of both matrix and membrane proteins. This cDNA encodes a hydrophilic protein (Pex19p) comprising 299 amino acids, with a prenylation motif, CAAX box, at the C terminus. Farnesylated Pex19p is partly, if not all, anchored in the peroxisomal membrane, exposing its N-terminal part to the cytosol. A stable transformant of ZP119 with HsPEX19 was morphologically and biochemically restored for peroxisome biogenesis. HsPEX19 expression also restored peroxisomal protein import in fibroblasts from a patient (PBDJ-01) with Zellweger syndrome of CG-J. This patient (PBDJ-01) possessed a homozygous, inactivating mutation: a 1-base insertion, A764, in a codon for Met255, resulted in a frameshift, inducing a 24-aa sequence entirely distinct from normal Pex19p. These results demonstrate that PEX19 is the causative gene for CG-J PBD and suggest that the C-terminal part, including the CAAX homology box, is required for the biological function of Pex19p. Moreover, Pex19p is apparently involved at the initial stage in peroxisome membrane assembly, before the import of matrix protein.

In vivo analysis of the 3′ untranslated region of the hepatitis C virus after in vitro mutagenesis of an infectious cDNA clone

Large sections of the 3′ untranslated region (UTR) of hepatitis C virus (HCV) were deleted from an infectious cDNA clone, and the RNA transcripts from seven deletion mutants were tested sequentially for infectivity in a chimpanzee. Mutants lacking all or part of the 3′ terminal conserved region or the poly(U–UC) region were unable to infect the chimpanzee, indicating that both regions are critical for infectivity in vivo. However, the third region, the variable region, was able to tolerate a deletion that destroyed the two putative stem–loop structures within this region. Mutant VR-24 containing a deletion of the proximal 24 nt of the variable region of the 3′ UTR was viable in the chimpanzee and seemed to replicate as well as the undeleted parent virus. The chimpanzee became viremic 1 week after inoculation with mutant VR-24, and the HCV genome titer increased over time during the early acute infection. Therefore, the poly(U–UC) region and the conserved region, but not the variable region, of the 3′ UTR seem to be critical for in vivo infectivity of HCV.

Identification of sample-specific sequences in mammalian cDNA and genomic DNA by the novel ligation-mediated subtraction (Limes)

The representational difference analysis (RDA) and other subtraction techniques are used to enrich sample-specific sequences by elimination of ubiquitous sequences existing in both the sample of interest (tester) and the subtraction partner (driver). While applying the RDA to genomic DNA of cutaneous lymphoma cells in order to identify tumor relevant alterations, we predominantly isolated repetitive sequences and artificial repeat-mediated fusion products of otherwise independent PCR fragments (PCR hybrids). Since these products severely interfered with the isolation of tester-specific fragments, we developed a considerably more robust and efficient approach, termed ligation-mediated subtraction (Limes). In first applications of Limes, genomic sequences and/or transcripts of genes involved in the regulation of transcription, such as transforming growth factor β stimulated clone 22 related gene (TSC-22R), cell death and cytokine production (caspase-1) or antigen presentation (HLA class II sequences), were found to be completely absent in a cutaneous lymphoma line. On the assumption that mutations in tumor-relevant genes can affect their transcription pattern, a protocol was developed and successfully applied that allows the identification of such sequences. Due to these results, Limes may substitute/supplement other subtraction/comparison techniques such as RDA or DNA microarray techniques in a variety of different research fields.

4-Coumarate:Coenzyme A Ligase in Hybrid Poplar1 : Properties of Native Enzymes, cDNA Cloning, and Analysis of Recombinant Enzymes

The enzyme 4-coumarate:coenzyme A ligase (4CL) is important in providing activated thioester substrates for phenylpropanoid natural product biosynthesis. We tested different hybrid poplar (Populus trichocarpa × Populus deltoides) tissues for the presence of 4CL isoforms by fast-protein liquid chromatography and detected a minimum of three 4CL isoforms. These isoforms shared similar hydroxycinnamic acid substrate-utilization profiles and were all inactive against sinapic acid, but instability of the native forms precluded extensive further analysis. 4CL cDNA clones were isolated and grouped into two major classes, the predicted amino acid sequences of which were 86% identical. Genomic Southern blots showed that the cDNA classes represent two poplar 4CL genes, and northern blots provided evidence for their differential expression. Recombinant enzymes corresponding to the two genes were expressed using a baculovirus system. The two recombinant proteins had substrate utilization profiles similar to each other and to the native poplar 4CL isoforms (4-coumaric acid > ferulic acid > caffeic acid; there was no conversion of sinapic acid), except that both had relatively high activity toward cinnamic acid. These results are discussed with respect to the role of 4CL in the partitioning of carbon in phenylpropanoid metabolism.

Parallel human genome analysis: microarray-based expression monitoring of 1000 genes.

Microarrays containing 1046 human cDNAs of unknown sequence were printed on glass with high-speed robotics. These 1.0-cm2 DNA "chips" were used to quantitatively monitor differential expression of the cognate human genes using a highly sensitive two-color hybridization assay. Array elements that displayed differential expression patterns under given experimental conditions were characterized by sequencing. The identification of known and novel heat shock and phorbol ester-regulated genes in human T cells demonstrates the sensitivity of the assay. Parallel gene analysis with microarrays provides a rapid and efficient method for large-scale human gene discovery.

Sequence analysis of a mannitol dehydrogenase cDNA from plants reveals a function for the pathogenesis-related protein ELI3.

Mannitol is the most abundant sugar alcohol in nature, occurring in bacteria, fungi, lichens, and many species of vascular plants. Celery (Apium graveolens L.), a plant that forms mannitol photosynthetically, has high photosynthetic rates thought to results from intrinsic differences in the biosynthesis of hexitols vs. sugars. Celery also exhibits high salt tolerance due to the function of mannitol as an osmoprotectant. A mannitol catabolic enzyme that oxidizes mannitol to mannose (mannitol dehydrogenase, MTD) has been identified. In celery plants, MTD activity and tissue mannitol concentration are inversely related. MTD provides the initial step by which translocated mannitol is committed to central metabolism and, by regulating mannitol pool size, is important in regulating salt tolerance at the cellular level. We have now isolated, sequenced, and characterized a Mtd cDNA from celery. Analyses showed that Mtd RNA was more abundant in cells grown on mannitol and less abundant in salt-stressed cells. A protein database search revealed that the previously described ELI3 pathogenesis-related proteins from parsley and Arabidopsis are MTDs. Treatment of celery cells with salicylic acid resulted in increased MTD activity and RNA. Increased MTD activity results in an increased ability to utilize mannitol. Among other effects, this may provide an additional source of carbon and energy for response to pathogen attack. These responses of the primary enzyme controlling mannitol pool size reflect the importance of mannitol metabolism in plant responses to divergent types of environmental stress.

A microarray study of post-mortem mRNA degradation in mouse brain tissue

Background: Although there is evidence that post-mortem interval (PMI) is not a major contributor to reduced overall RNA integrity, it may differentially affect a subgroup of gene transcripts that are susceptible to PMI-related degradation. This would particularly have ramifications for microarray studies that include a broad spectrum of genes. Method: Brain tissue was removed from adult mice at 0, 6, 12, 18, 24,36 and 48 h post-mortem. RNA transcript abundance was measured by hybridising RNA from the zero time point with test RNA from each PMI time point, and differential gene expression was assessed using cDNA microarrays. Sequence and ontological analyses were performed on the group of RNA transcripts showing greater than two-fold reduction. Results: Increasing PMI was associated with decreased tissue pH and increased RNA degradation as indexed by 28S/18S ribosomal RNA ratio. Approximately 12% of mRNAs detected on the arrays displayed more than a two-fold decrease in abundance by 48 It post-mortem. An analysis of nucleotide composition provided evidence that transcripts with the AUUUA motif in the 3' untranslated region (3'UTR) were more susceptible to PMI-related RNA degradation, compared to transcripts not carrying the 3'UTR AUUUA motif. Consistent with this finding, ontological analysis showed transcription factors and elements to be over-represented in the group of transcripts susceptible to degradation. Conclusion: A subgroup of mammalian mRNA transcripts are particularly susceptible to PMI-related degradation, and as a group, they are more likely to carry the YUTR AUUUA motif. PMI should be controlled for in human and animal model post-mortem brain studies, particularly those including a broad spectrum of mRNA transcripts. (c) 2005 Elsevier B.V. All rights reserved.

Cloning and sequence analysis of pituitary prolactin cDNA from the northern brown bandicoot (Isoodon macrourus)

The nuclectide sequence for pituitary prolactin cDNA from the marsupial bandicoot (Isoodon macrourus) was determined by reverse transcription-polymerase chain reaction and 5'/3' rapid amplification of cDNA ends. The deduced amino acid sequence showed high sequence identity with brushtail possum prolactin (95%) and all of the expected structural features of a quadruped prolactin. A prolactin gene tree was constructed and rates of evolution calculated for bandicoot, possum, opossum and several mammalian and non-mammalian prolactins. Bootstrap analysis provided strong support for marsupials as a sister group with eutherian mammals and weak support for opossum and bandicoot as an independent grouping from the brushtail possum. The rates of molecular evolution for marsupial prolactins were comparable to the slow rate seen in the majority of quadruped prolactins that have been sequenced. (c) 2005 Elsevier Inc. All rights reserved.

Expressed sequence tag analysis of genes expressed during development of the tropical abalone Haliotis asinina

The tropical abalone. Haliotis asinina. is,in ideal species to investigate the molecular mechanisms that control development. growth, reproduction and shell formation in all cultured haliotids. Here we describe the analysis of 232 expressed sequence tags (EST) obtained front a developmental H. asinina cDNA library intended for future microarray studies. From this data set we identified 183 unique gene Clusters. Of these, 90 clusters showed significant homology with sequences lodged in GenBank, ranging in function from general housekeeping to signal transduction, gene regulation and cell-cell communication. Seventy-one clusters possessed completely novel ORFs greater than 50 codons in length, highlighting the paucity of sequence data from molluscs and other lophotrochozoans. This study of developmental gene expression in H. asinina provides the foundation for further detailed analyses of abalone growth, development and reproduction.

Transcriptome analysis of Neotyphodium and Epichloë grass endophytes

Large-scale gene discovery has been performed for the grass fungal endophytes Neotyphodium coenophialum, Neotyphodium lolii, and Epichloe festucae. The resulting sequences have been annotated by comparison with public DNA and protein sequence databases and using intermediate gene ontology annotation tools. Endophyte sequences have also been analysed for the presence of simple sequence repeat and single nucleotide polymorphism molecular genetic markers. Sequences and annotation are maintained within a MySQL database that may be queried using a custom web interface. Two cDNA-based microarrays have been generated from this genome resource, They permit the interrogation of 3806 Neotyphodium genes (Nchip (TM) rnicroarray), and 4195 Neotyphodium and 920 Epichloe genes (EndoChip (TM) microarray), respectively. These microarrays provide tools for high-throughput transcriptome analysis, including genome-specific gene expression studies, profiling of novel endophyte genes, and investigation of the host grass-symbiont interaction. Comparative transcriptome analysis in Neotyphodium and Epichloe was performed. (c) 2006 Elsevier

On the simultaneous use of clinical and microarray expression data in the cluster analysis of tissue samples

This paper considers a model-based approach to the clustering of tissue samples of a very large number of genes from microarray experiments. It is a nonstandard problem in parametric cluster analysis because the dimension of the feature space (the number of genes) is typically much greater than the number of tissues. Frequently in practice, there are also clinical data available on those cases on which the tissue samples have been obtained. Here we investigate how to use the clinical data in conjunction with the microarray gene expression data to cluster the tissue samples. We propose two mixture model-based approaches in which the number of components in the mixture model corresponds to the number of clusters to be imposed on the tissue samples. One approach specifies the components of the mixture model to be the conditional distributions of the microarray data given the clinical data with the mixing proportions also conditioned on the latter data. Another takes the components of the mixture model to represent the joint distributions of the clinical and microarray data. The approaches are demonstrated on some breast cancer data, as studied recently in van't Veer et al. (2002).

Microarray and proteomic analysis of the human alcoholic brain

The superior frontal cortex (SFC) is selectively damaged in chronic alcohol abuse, with localized neuronal loss and tissue atrophy. Regions such as motor cortex show little neuronal loss except in severe co-morbidity (liver cirrhosis or WKS). Altered gene expression was found in microarray comparisons of alcoholic and control SFC samples [1]. We used Western blots and proteomic analysis to identify the proteins that also show differential expression. Tissue was obtained at autopsy under informed, written consent from uncomplicated alcoholics and age- and sex-matched controls. Alcoholics had consumed 80 g ethanol/day chronically (often, 200 g/day for 20 y). Controls either abstained or were social drinkers ( 20 g/day). All subjects had pathological confirmation of liver and brain diagnosis; none had been polydrug abusers. Samples were homogenized in water and clarified by brief centrifugation (1000g, 3 min) before storage at –80°C. For proteomics the thawed suspensions were centrifuged (15000g, 50 min) to prepare soluble fractions. Aliquots were pooled from SFC samples from the 5 chronic alcoholics and 5 matched controls used in the previous microarray study [1]. 2-Dimensional electrophoresis was performed in triplicate using 18 cm format pH 4–7 and pH 6–11 immobilized pH gradients for firstdimension isoelectric focusing. Following second-dimension SDS-PAGE the proteins were fluorescently stained and the images collected by densitometry. 182 proteins differed by 2-fold between cases and controls. 141 showed lower expression in alcoholics, 33 higher, and 8 were new or had disappeared. To date 63 proteins have been identified using MALDI-MS and MS-MS. Western blots were performed on uncentrifuged individual samples from 76 subjects (controls, uncomplicated alcoholics and cirrhotic alcoholics). A common standard was run on every gel. After transfer, immunolabeling, and densitometry, the intensities of the unknown bands were compared to those of the standards. We focused on proteins from transcripts that showed clear differences in a series of microarray studies, classified into common sets including Regulators of G-protein Signaling and Myelin-associated proteins. The preponderantly lower level of differentially expressed proteins in alcoholics parallels the microarray mRNA analysis in the same samples. We found that mRNA and protein expression do not frequently correspond; this may help identify pathogenic processes acting at the level of transcription, translation, or post-translationally.

A Comparative Analysis of Predictive Learning Algorithms on High-Dimensional Microarray Cancer Data

This research evaluates pattern recognition techniques on a subclass of big data where the dimensionality of the input space (p) is much larger than the number of observations (n). Specifically, we evaluate massive gene expression microarray cancer data where the ratio κ is less than one. We explore the statistical and computational challenges inherent in these high dimensional low sample size (HDLSS) problems and present statistical machine learning methods used to tackle and circumvent these difficulties. Regularization and kernel algorithms were explored in this research using seven datasets where κ < 1. These techniques require special attention to tuning necessitating several extensions of cross-validation to be investigated to support better predictive performance. While no single algorithm was universally the best predictor, the regularization technique produced lower test errors in five of the seven datasets studied.