HykGene: a hybrid approach for selecting marker genes for phenotype classification using microarray gene expression data

MOTIVATION: Recent studies have shown that microarray gene expression data are useful for phenotype classification of many diseases. A major problem in this classification is that the number of features (genes) greatly exceeds the number of instances (tissue samples). It has been shown that selecting a small set of informative genes can lead to improved classification accuracy. Many approaches have been proposed for this gene selection problem. Most of the previous gene ranking methods typically select 50-200 top-ranked genes and these genes are often highly correlated. Our goal is to select a small set of non-redundant marker genes that are most relevant for the classification task. RESULTS: To achieve this goal, we developed a novel hybrid approach that combines gene ranking and clustering analysis. In this approach, we first applied feature filtering algorithms to select a set of top-ranked genes, and then applied hierarchical clustering on these genes to generate a dendrogram. Finally, the dendrogram was analyzed by a sweep-line algorithm and marker genes are selected by collapsing dense clusters. Empirical study using three public datasets shows that our approach is capable of selecting relatively few marker genes while offering the same or better leave-one-out cross-validation accuracy compared with approaches that use top-ranked genes directly for classification. AVAILABILITY: The HykGene software is freely available at http://www.cs.dartmouth.edu/~wyh/software.htm CONTACT: wyh@cs.dartmouth.edu SUPPLEMENTARY INFORMATION: Supplementary material is available from http://www.cs.dartmouth.edu/~wyh/hykgene/supplement/index.htm.     

URA3 as a selectable marker for disruption and virulence assessment of Candida albicans genes

The ability to generate isogenic sets of strains with mutations in a gene of interest but not in other genes by repeated use of the URA3 marker (Ura-blaster methodology) has advanced our understanding of the relationships between gene structure and function in Candida albicans. Common applications of Ura-blaster technology result in different genomic positions for the URA3 gene in strains complemented for the gene of interest compared with mutant strains. Studies using animal models of systemic candidiasis pointed to possible differences in URA3 gene expression, depending on its genomic location, which confounded interpretation of the role of the gene of interest in lethality. Positional effects on URA3 expression can be avoided by placement at a common locus in all strains used for comparison.     

Novel genes underlying beta cell survival in metabolic stress

Relative insulin deficiency, in response to increased metabolic demand (obesity, genetic insulin resistance, pregnancy and aging) lead to Type2 diabetes. Susceptibility of the type 2 diabetes has a genetic basis, as a subset of people with risk factors (obesity, Insulin Resistance, pregnancy), develop Type2 Diabetes. We aimed to identify ‘cluster’ of overexpressed genes, underlying increased beta cell survival in diabetes resistant C57BL/6J ob/ob mice (compared to diabetes susceptible BTBR ob/ob mice). We used ‘consensus’ overexpression status to identify ‘cluster’ of 11 genes consisting of Aldh18a1, Rfc4, Dynlt3, Prom1, H13, Psen1, Ssr4, Dad1, Anpep, Fam111a and Plk1. Information (biological processes, molecular functions, cellular components, protein-protein interactions/associations, gene deletion/knockout/inhibition studies) of all the genes in ‘cluster’ were collected by text mining using different literature search tools, gene information databases and protein-protein interaction databases. Beta cell specific function of these genes were also inferred using meta analysis tool of Beta Cell Biology Consortium, by studying the expression pattern of these genes in microarray studies related to beta-cell stimulation/injury, pancreas development and growth and cell differentiation. In the ‘clusters’, 6 genes (Dad1, Psen1, Ssr4, Rfc4, H13, Plk1) have a role in cell survival. Only Psen1 was previously identified to have role in successful beta cell compensation. We advocate these genes to be potentially involved in successful beta cell compensation and prevent T2D in humans, by conferring protection against diabetogenic insults.     

Genomic assessment of human cumulus cell marker genes as predictors of oocyte developmental competence: impact of various experimental factors

Background

Single embryo transfer (SET) is the most successful way to reduce the frequency of multiple pregnancies following in vitro fertilisation. However, selecting the embryo for SET with the highest chances of pregnancy remains a difficult challenge since morphological and kinetics criteria provide poor prediction of both developmental and implantation ability. Partly through the expression of specific genes, the oocyte-cumulus interaction helps the oocyte to acquire its developmental competence. Our aim was therefore to identify at the level of cumulus cells (CCs) genes related to oocyte developmental competence.

Methodology/Principal Findings

197 individual CCs were collected from 106 patients undergoing an intra-cytoplasmic sperm injection procedure. Gene expression of CCs was studied using microarray according to the nuclear maturity of the oocyte (immature vs. mature oocyte) and to the developmental competence of the oocyte (ability to reach the blastocyst stage after fertilisation). Microarray study was followed by a meta-analysis of the behaviour of these genes in other datasets available in Gene Expression Omnibus which showed the consistency of this list of genes. Finally, 8 genes were selected according to oocyte developmental competence from the 308 differentially expressed genes (p<0.0001) for further validation by quantitative PCR (qPCR). Three of these 8 selected genes were validated as potential biomarkers (PLIN2, RGS2 and ANG). Experimental factors such as inter-patient and qPCR series variability were then assessed using the Generalised Linear Mixed Model procedure, and only the expression level of RGS2 was confirmed to be related to oocyte developmental competence. The link between biomarkers and pregnancy was finally evaluated and level of RGS2 expression was also correlated with clinical pregnancy.

Conclusion/Significance

RGS2, known as a regulator of G protein signalling, was the only gene among our 8 selected candidates biomarkers of oocyte competence to cover many factors of variability, including inter-patient factors and experimental conditions.     

Coding sequence polymorphisms among V beta T cell receptor genes

The four V beta gene segments, V beta 1, 3.1, 6, and 10, which previously have been shown by RFLP analyses to differ between TCR V beta a and V beta b haplotypes, were cloned and sequenced from V beta a SWR mice and compared with V beta b strains to define coding sequence polymorphisms distinguishing these haplotypes. V beta 3.1 and 6 alleles differed between strains by a single amino acid, whereas V beta 1 and 10 alleles differed by 4 and 6 amino acids, respectively. The overall interhaplotypic V beta polymorphisms appeared to be limited when based upon compilation of this information and previously published V beta sequences. One application of these data was to attempt to elucidate the molecular basis underlying the recently reported allele-specific autoimmune disease, collagen-induced arthritis. Based on the present structural data and the additional evidence, contrary to what was suggested, the V beta 6 gene does not appear to be the sole participant in anticollagen responses.     

Equine marker genes: Polymorphism for plasminogen

Polymorphism for two autosomal alleles of equine plasminogen, PLG ¹ and PLG ², was demonstrated in plasma by isoelectric focusing and immunofixation, with a goat anti-human plasminogen antibody. The frequency of PLC ² was 0.16 in 150 Standardbreds, 0.20 in 96 Thoroughbreds, and 0.39 in 32 Shetland ponies. No evidence for linkage of PLG with any of 13 marker loci was found.     

Equine marker genes: polymorphism for plasminogen

Polymorphism for two autosomal alleles of equine plasminogen, PLG1 and PLG2, was demonstrated in plasma by isoelectric focusing and immunofixation, with a goat anti-human plasminogen antibody. The frequency of PLG2 was 0.16 in 150 Standardbreds, 0.20 in 96 Thoroughbreds, and 0.39 in 32 Shetland ponies. No evidence for linkage of PLG with any of 13 marker loci was found.     

Lifespan regulation by evolutionarily conserved genes essential for viability

Evolutionarily conserved mechanisms that control aging are predicted to have prereproductive functions in order to be subject to natural selection. Genes that are essential for growth and development are highly conserved in evolution, but their role in longevity has not previously been assessed. We screened 2,700 genes essential for Caenorhabditis elegans development and identified 64 genes that extend lifespan when inactivated postdevelopmentally. These candidate lifespan regulators are highly conserved from yeast to humans. Classification of the candidate lifespan regulators into functional groups identified the expected insulin and metabolic pathways but also revealed enrichment for translation, RNA, and chromatin factors. Many of these essential gene inactivations extend lifespan as much as the strongest known regulators of aging. Early gene inactivations of these essential genes caused growth arrest at larval stages, and some of these arrested animals live much longer than wild-type adults. daf-16 is required for the enhanced survival of arrested larvae, suggesting that the increased longevity is a physiological response to the essential gene inactivation. These results suggest that insulin-signaling pathways play a role in regulation of aging at any stage in life.     

Plant mutant tubulin genes as marker selective genes for genetic engineering

The approaches for new marker genes usage in selection of transformed plant cells, which are based on using mutant tubulin genes from natural plant biotypes and, in perspective, from induced plant mutants have been considered. The results of investigations of plant (biotypes, mutants) resistance to herbicides with antimicrotubular mode of action on molecular and cellular levels have been summarized. The reports dealing with study the transferring and the expression of mutant tubulin genes conferring resistance to amiprophosmethyl (phosphorothioamidate herbicide) and to trifluralin (dinitroaniline herbicide) from corresponding N. plumbaginifolia mutants into related and remote plant species by somatic hybridisation methods have been analyzed. The results of experiments on monocotyledonous and dicotyledonous. plant transformation by mutant alpha-tubulin gene conferring resistance to dinitroanilines are described to test the possibility of its using as a marker gene with obtaining, at the same time, a dinitroaniline-resistant plants.     

Mutant genes of plant tubulins as selective marker genes for genetic engineering

This review describes different approaches to employment of new marker genes in selection of transformed plant cells, which are based on the use of mutant tubulin genes from natural plant biotypes and, in prospect, induced plant mutants. The results of studies of plant (biotypes, mutants) resistance to herbicides with antimicrotubular mode of action at molecular and cellular levels were summarized. The reports on the transfer and expression of mutant tubulin genes conferring resistance to amiprophosmethyl (phosphorothioamidate herbicide) and trifluralin (dinitroaniline herbicide) from corresponding Nicotiana plumbaginifolia mutants in related and remote plant species by somatic hybridization methods were analyzed. The results of experiments on transformation of monocotyledonous and dicotyledonous plants by mutant α-tubulin gene conferring resistance to dinitroanilines are described to test the possibility of its use as a marker gene and simultaneously obtaining dinitroaniline-resistant plants.     

Clusters of genes encoding mouse transplantation antigens

We constructed a cosmid library from BALB/c mouse sperm DNA and isolated 64 cosmid clones with cDNA probes for transplantation antigens (class I molecules). Of these clones, 54 mapped into 13 gene clusters containing 36 distinct class I genes and encompassing 837 kilobases of DNA. One gene cluster mapped to the L region and a second cluster with seven genes to the Qa-2,3 region of the major histocompatibility complex. Restriction map and Southern blot analyses suggest that there are subgroups of class I genes. Using a 5' flanking sequence of the L gene as a hybridization probe, we show the L gene to be present in mouse strains expressing this antigen but deleted or mutated in strains failing to express it. Our data suggest that gene duplication and deletion presumably by homologous but unequal crossing-over has altered the size and organization of the class I clusters in different mouse strains and probably is an important mechanism for generating polymorphism in these genes. Analysis of the 36 class I genes with cDNA probes specific for the 5' and 3' ends shows that the exon encoding the third external domain is far more conserved than those encoding the first and second external domains of the transplantation antigen. These differences in variability have interesting functional implications.     

Clusters of genes for the biosynthesis of antibiotics: Regulatory genes and overproduction of pharmaceuticals

Summary In the last decade numerous genes involved in the biosynthesis of antibiotics, pigments, herbicides and other secondary metabolites have been cloned. The genes involved in the biosynthesis of penicillin, cephalosporin and cephamycins are organized in clusters as occurs also with the biosynthetic genes of other antibiotics and secondary metabolites (see review by Martín and Liras [65]). We have cloned genes involved in the biosynthesis of -lactam antibiotics from five different -lactam producing organisms both eucaryotic (Penicillium chrysogenum, Cephalosporium acremonium (syn.Acremonium chrysogenum) Aspergillus nidulans) and procaryotic (Nocardia lactamdurans, Streptomyces clavuligerus). InP. chrysogenum andA. nidulans the organization of thepcbAB,pcbC andpenDE genes for ACV synthetase, IPN synthase and IPN acyltransferase showed a similar arrangement. InA. chrysogenum two different clusters of genes have been cloned. The cluster of early genes encodes ACV synthetase and IPN synthase, whereas the cluster of late genes encodes deacetoxycephalosporin C synthetase/hydroxylase and deacetylcephalosporin C acetyltransferase. InN. lactamdurans andS. clavuligerus a cluster of early cephamycin genes has been fully characterized. It includes thelat (for lysine-6-aminotransferase),pcbAB (for ACV synthase) andpcbC (for IPN synthase) genes. Pathway-specific regulatory genes which act in a positive (or negative) form are associated with clusters of genes involved in antibiotic biosynthesis. In addition, widely acting positive regulatory elements exert a pleiotropic control on secondary metabolism and differentiation of antibiotic producing microorganisms.The application of recombinant DNA techniques will contribute significantly to the improvement of fermentation organisms.     

Rearranged beta T cell receptor genes in a helper T cell clone specific for lysozyme: no correlation between V beta and MHC restriction

The helper T cell clone 3H.25 is specific for hen egg white lysozyme and the class II MHC molecule I-Ab. This TH cell has three rearrangements in the beta-chain gene family-a V beta-D beta-J beta 1 and a D beta 2-J beta 2 rearrangement on one homolog and a D beta 1-J beta 2 rearrangement on the other. These observations demonstrate that this functional T lymphocyte expresses only a single V beta gene segment and, accordingly, exhibits allelic exclusion of beta-chain gene expression. The rearranged 3H.25 V beta gene segment is the same as that expressed in a T helper cell specific for cytochrome c and an I-Ek MHC molecule. Thus, there is no simple correlation between the V beta gene segment and antigen specificity or MHC restriction.     

Advances in selectable marker genes for plant transformation

Plant transformation systems for creating transgenics require separate process for introducing cloned DNA into living plant cells. Identification or selection of those cells that have integrated DNA into appropriate plant genome is a vital step to regenerate fully developed plants from the transformed cells. Selectable marker genes are pivotal for the development of plant transformation technologies because marker genes allow researchers to identify or isolate the cells that are expressing the cloned DNA, to monitor and select the transformed progeny. As only a very small portion of cells are transformed in most experiments, the chances of recovering transgenic lines without selection are usually low. Since the selectable marker gene is expected to function in a range of cell types it is usually constructed as a chimeric gene using regulatory sequences that ensure constitutive expression throughout the plant. Advent of recombinant DNA technology and progress in plant molecular biology had led to a desire to introduce several genes into single transgenic plant line, necessitating the development of various types of selectable markers. This review article describes the developments made in the recent past on plant transformation systems using different selection methods adding a note on their importance as marker genes in transgenic crop plants.     

Positive selectable marker genes for routine plant transformation

Summary Plant genetic transformation technologies rely upon the selection and recovery of transformed cells. Selectable marker genes used so far have been either antibiotic resistance genes or herbicide tolerance genes. There is a need to apply alternative principles of selection, as more transgenic traits have to be incorporated into a transgenic crop and because of concern that the use of conventional marker genes may pose a threat to humans and the environment. New classes of marker genes are now available, conferring metabolic advantage of the transgenic cells over the non-transformed cells. The new selection systems, as described in this review, are being used with success and superior performance over the traditional marker systems.