PubMed Health⌕ Search

Biomedical subjects

E Boerwinkle

Publications and source records attributed to E Boerwinkle.

At least 19 recordsLinked to original sources

The contribution of individual and pairwise combinations of SNPs in the APOA1 and APOC3 genes to interindividual HDL-C variability.

Apolipoproteins (apo) A-I and C-III are components of high-density lipoprotein-cholesterol (HDL-C), a quantitative trait negatively correlated with risk of cardiovascular disease (CVD). We analyzed the contribution of individual and pairwise combinations of single nucleotide polymorphisms (SNPs) in the APOA1/APOC3 genes to HDL-C variability to evaluate (1) consistency of published single-SNP studies with our single-SNP analyses; (2) consistency of single-SNP and two-SNP phenotype-genotype relationships across race-, gender-, and geographical location-dependent contexts; and (3) the contribution of single SNPs and pairs of SNPs to variability beyond that explained by plasma apo A-I concentration. We analyzed 45 SNPs in 3,831 young African-American (N=1,858) and European-American (N=1,973) females and males ascertained by the Coronary Artery Risk Development in Young Adults (CARDIA) study. We found three SNPs that significantly impact HDL-C variability in both the literature and the CARDIA sample. Single-SNP analyses identified only one of five significant HDL-C SNP genotype relationships in the CARDIA study that was consistent across all race-, gender-, and geographical location-dependent contexts. The other four were consistent across geographical locations for a particular race-gender context. The portion of total phenotypic variance explained by single-SNP genotypes and genotypes defined by pairs of SNPs was less than 3%, an amount that is miniscule compared to the contribution explained by variability in plasma apo A-I concentration. Our findings illustrate the impact of context-dependence on SNP selection for prediction of CVD risk factor variability.

Adolescent↗

Antihypertensive therapy, the alpha-adducin polymorphism, and cardiovascular disease in high-risk hypertensive persons: the Genetics of Hypertension-Associated Treatment Study.

In a double-blind, outcome trial conducted in hypertensive patients randomized to chlorthalidone (C), amlodipine (A), lisinopril (L), or doxazosin (D), the alpha-adducin Gly460Trp polymorphism was typed (n=36 913). Mean follow-up was 4.9 years. Relative risks (RRs) of chlorthalidone versus other treatments were compared between genotypes (Gly/Gly+Gly/Trp versus Trp/Trp). Primary outcome was coronary heart disease (CHD). Coronary heart disease incidence did not differ among treatments or genotypes nor was there any interaction between treatment and genotype (P=0.660). Subgroup analyses indicated that Trp allele carriers had greater CHD risk with C versus A+L in women (RR=1.31) but not men (RR=0.91) with no RR gender differences for non-carriers (gender-gene-treatment interaction, P=0.002). The alpha-adducin gene is not an important modifier of antihypertensive treatment on cardiovascular risk, but women Trp allele carriers may have increased CHD risk if treated with C versus A or L. This must be confirmed to have implications for hypertension treatment.

Aged↗

Influence of lipoprotein lipase gene Ser447Stop and beta1-adrenergic receptor gene Arg389Gly polymorphisms and their interaction on obesity from childhood to adulthood: the Bogalusa Heart Study.

OBJECTIVE: To investigate the influence of lipoprotein lipase (LPL) Ser447Stop and beta1-adrenergic receptor (ADRB1) Arg389Gly gene polymorphisms, individually and in combination, on obesity from childhood to adulthood. DESIGN AND SUBJECTS: A community-based cohort of 1331 subjects (30% black and 70% white subjects) was followed over an average period of 23 years from childhood (age range: 4-17 years) to adulthood (age range:18-44 years). MEASUREMENT: Body mass index (BMI, kg/m2) and LPL Ser447Stop and the ADRB1 Arg389Gly genotypes. RESULTS: The frequency of the ADRB1 Gly389 allele was 0.25 in white subjects vs 0.39 in black subjects (P < 0.001); 0.08 vs 0.05 (P = 0.280) for the LPL Stop447 allele. There was no association between the LPL Stop447 allele and BMI among white and black subjects either in childhood and adulthood levels or annual change from childhood to adulthood. The ADRB1 Gly389 allele was associated with lower BMI only in black adults (P = 0.017). Further, the interaction effect of the LPL Stop447 allele and ADRB1 Gly389 allele on adult BMI or its annual change was significant in white subjects and in the total sample (P = 0.03-0.006). Childhood values tended to show a similar trend. Having both ADRB1 Gly389 allele and LPL Stop447 allele was associated with 71% (95% confidence interval: 26-89%) less odds for developing obesity from childhood to adulthood after adjusting for age, race, sex, and childhood BMI. CONCLUSION: While Gly389 allele of the ADRB1 gene lowers obesity in black subjects, this allele in conjunction with Stop447 allele of the LPL gene lowers obesity in adults and attenuates the development of obesity from childhood to adulthood. These findings underscore the importance of gene-gene interaction in the assessment of genetic influences on complex traits such as obesity.

Adolescent↗

Association between patterns of nucleotide variation across the three fibrinogen genes and plasma fibrinogen levels: the Coronary Artery Risk Development in Young Adults (CARDIA) study.

BACKGROUND: Previous genotype-phenotype association studies of fibrinogen have been limited by incomplete knowledge of genomic sequence variation within and between major ethnic groups in FGB, FGA, and FGG. METHODS: We characterized the linkage disequilibrium patterns and haplotype structure across the human fibrinogen gene locus in European- and African-American populations. We analyzed the association between common polymorphisms in the fibrinogen genes and circulating levels of both 'functional' fibrinogen (measured by the Clauss clotting rate method) and total fibrinogen (measured by immunonephelometry) in a large, multi-center, bi-racial cohort of young US adults. RESULTS: A common haplotype tagged by the A minor allele of the well-studied FGB-455 G/A promoter polymorphism (FGB 1437) was confirmed to be strongly associated with increased plasma fibrinogen levels. Two non-coding variants specific to African-American chromosomes, FGA 3845 A and FGG 5729 G, were each associated with lower plasma fibrinogen levels. In European-Americans, a common haplotype tagged by FGA Thr312Ala and several other variant alleles across the fibrinogen gene locus was strongly associated with decreased fibrinogen levels as measured by functional assay, but not by immunoassay. Overall, common polymorphisms within the three fibrinogen genes explain < 2% of the variability in plasma fibrinogen concentration. CONCLUSIONS: In young adults, fibrinogen multi-locus genotypes are associated with plasma fibrinogen levels. The specific single nucleotide polymorphism and haplotype patterns for these associations differ according to population and also according to phenotypic assay. It is likely that a substantial proportion of the heritable component of plasma fibrinogen concentration is due to genetic variation outside the three fibrinogen genes.

Adolescent↗

APOE genotype and cognitive decline in a middle-aged cohort.

BACKGROUND: Most longitudinal studies of nondemented persons have reported greater cognitive decline among APOE epsilon4 carriers vs noncarriers. However, most studies involved elderly samples (aged 65+) and were not large enough to examine the three APOE alleles separately. METHODS: Change in cognitive function was examined over a 6-year period using three neuropsychological tests among four APOE genotype groups (epsilon2/2 + epsilon2/3, epsilon3/3 (referent), epsilon4/2 + epsilon4/3, epsilon4/4). The population-based sample included 1,693 African Americans and 6,202 Caucasians initially ages 47 to 68. RESULTS: There was increasingly greater cognitive decline from the epsilon2 group to the epsilon4/4 group in Caucasians for two of the three tests. The combination of APOE epsilon4 with hypercholesterolemia or diabetes showed the greatest cognitive decline. Among African Americans, only the test measuring psychomotor speed showed associations with APOE genotype. CONCLUSIONS: APOE epsilon4 is associated with greater cognitive decline in middle-aged Caucasian individuals, with a reduced decline among epsilon2 carriers. This suggests that the processes by which APOE genotype mediates dementia risk are operative well in advance of overt dementia.

Black or African American↗

Linkage analysis of plasma ApoE in three ethnic groups: multiple genes with context-dependent effects.

We performed variance component-based linkage analysis in four samples (two of non-Hispanic European-Americans from Rochester, MN; African-Americans from Jackson, MS; and Mexican-Americans from Starr County, TX) to identify chromosomal regions containing genes influencing plasma apolipoprotein E (apoE) levels. The APOE gene region on chromosome (chr) 19 was identified with a LOD > or = 2.00 in both samples from Rochester and the sample from Jackson. Adjustment of apoE levels for differences among means of genotypes defined by the APOE epsilon2/3/4 alleles reduced evidence of linkage, indicating that the APOE gene was responsible for the majority of the linkage signal. In stratified linkage analyses, there was a LOD of 1.70 in the Starr County sibships with average total cholesterol (TC) above the median level for all sibships in that population. Adjustment for APOE genotype did not remove this LOD score, suggesting a second gene in this region may influence apoE variation. Evidence of linkage (LOD= 3.32) on chr 17 was observed in the Starr County sibships with average TC below the median. Inter-individual variation in plasma apoE level may be influenced by variations in the structural gene, and at least one other gene whose effects differ among populations and are dependent on the influence of unmeasured genetic and environmental factors indexed by correlated measures of lipid metabolism.

Apolipoproteins E↗

An autosomal genomic scan for loci linked to type 2 diabetes in northern Han Chinese.

We report the results of a genome-wide scan conducted in 219 individuals from 34 large multiplex nuclear pedigrees from the northern Han Chinese population at an average resolution of about 10 cM. Nonparametric two-point and multipoint linkage analyses were performed to detect evidence of linkage with type 2 diabetes in this study. On chromosome 1 four regions showed evidence of linkage with type 2 diabetes in northern Han Chinese. Of these regions a marker D1S193 (73 cM) showed evidence of linkage (two-point nonparametric linkage 2.409), and another region (around 190 cM) was a replication of several other studies performed in different ethnic populations. Evidences of linkage have been confirmed by typing additional markers (average distance 1-5 cM) flanking these two positive regions on chromosome 1. We also found indication of linkage with type 2 diabetes on chromosomes 2, 10, 12, 18, 20, and 22 by two-point linkage analyses.

Adult↗

An autosomal genome scan for loci influencing longitudinal burden of body mass index from childhood to young adulthood in white sibships: The Bogalusa Heart Study.

OBJECTIVE: To examine genetic loci linked to a long-term burden and trend of obesity traits, such as body mass index (BMI), from childhood to adulthood. DESIGN: : Longitudinal study using serial measurements of BMI from childhood. SUBJECTS: A total of 782 unselected white siblings (representing 521 full and 39 half sib-pairs) from 342 families enrolled in the Bogalusa Heart Study. MEASUREMENTS: A total of 357 microsatellite markers with an average spacing of 9.0 cM spanning the 22 autosomal chromosomes were typed. A quadratic growth curve was developed using a random effects model based on serial measurements of BMI from childhood to adulthood. The serial changes in BMI were measured in terms of long-term burden (area under the curve (AUC) divided by follow-up years) and the long-term trend (incremental AUC, calculated as total AUC-baseline AUC). RESULTS: Heritability estimates of long-term measures were 0.78 for total AUC and 0.43 for incremental AUC. In a variance-component-based multipoint linkage analysis with SOLAR, linkage to the long-term measures of BMI was observed on chromosomes 1, 5, 7, 12, 13 and 18. For total AUC, LOD scores were 3.0 at 110 cM on chromosome 12, 2.9 at 26 cM and 2.4 at 52 cM on chromosome 7, and 2.2 at 126 cM on chromosome 5. For incremental AUC, LOD scores were 2.9 at 26 cM, 2.1 at 97 cM and 2.3 at 110 cM on chromosome 12, 2.2 at 69 cM on chromosome 7, 2.2 at 91 cM and 2.5 at 150 cM on chromosome 1, 2.0 at 119 cM on chromosome 5, 2.0 at 54 cM on chromosome 13 and 2.0 at 7 cM on chromosome 18. Several important obesity-related candidate genes are located in the regions or near the markers showing positive linkage. CONCLUSION: Linkage evidence found in this study indicates that regions on these chromosomes might harbor genetic loci that affect the propensity to develop obesity from childhood.

Adolescent↗

Multilocus effects of the renin-angiotensin-aldosterone system genes on blood pressure response to a thiazide diuretic.

BACKGROUND: The renin-angiotensin-aldosterone (RAA) system regulates blood pressure (BP) levels and influences responses to antihypertensive medications. Variation in RAA system genes has been reported to influence interindividual differences in BP levels and the occurrence of hypertension (HTN). METHODS: We evaluated the relationship between variation in genes of the RAA system and interindividual differences in BP response to a thiazide diuretic. Analyses were carried out in a race- and gender-specific manner in 255 unrelated hypertensive African-Americans (125 men and 130 women) and 246 unrelated hypertensive non-Hispanic Whites (133 men and 113 women). RESULTS: The angiotensin II receptor (AT(1)R) A1166C and angiotensinogen G-6A polymorphisms had a significant effect on systolic BP response to the diuretic in African-American women. Multilocus analyses indicated that the effects of these genes combined additively to influence response. Results of a permutation test to adjust for multiple comparisons and the possible nonindependence among genotypes remained significant at the P=0.003 level. CONCLUSIONS: Among African-American women, particular gene variations in the RAA system have additive effects on BP response to a thiazide diuretic.

Adult↗

Evidence for non-additive influence of single nucleotide polymorphisms within the apolipoprotein E gene.

We analyzed 13 single nucleotide polymorphisms (SNPs) within the apolipoprotein E (APOE) gene, to identify pairs of SNPs that interact in a non-additive manner to influence genotypic mean levels of the ApoE protein in blood. An overparameterized general linear model of two-SNP genotype means was applied to data from 456 female and 398 male unrelated European Americans from Rochester, MN, USA. We found statistically significant evidence for non-additivity between SNPs within the male sample, but not within the female sample. We observed nine pairs of SNPs with evidence of non-additivity at the alpha=0.05 level of statistical significance within the male sample, when approximately three were expected by chance. Five of the nine pairs involved three SNPs (560, 624 and 1163) that did not have a statistically significant influence when considered separately in a single-site analysis. Three of the nine pairs involving four SNPs (832, 1998, 3937 and 4951) showed significant evidence for non-additivity in at least one of two other male samples from Jackson, MS, USA and North Karelia, Finland. Although all four of these SNPs had a statistically significant influence in Rochester when considered separately, only SNP 3937 gave a significant result in the other male samples. The four SNPs are located in the promoter, intronic and exonic regions, and 3' to the polyadenylation signal in the APOE gene. Our study suggests that analyses that only consider SNPs located in exons and ignore contexts such as those indexed by gender and population, and disregard non-additivity of SNP effects, may inappropriately model the contribution of a gene to the genetic architecture of a trait that has a complex multifactorial etiology.

Apolipoproteins E↗

Power of a simplified multivariate test for genetic linkage.

In this paper we compare the power of the multivariate Haseman-Elston (MHE) test proposed earlier by Amos et al. (1990) and a computationally rapid new version of the multivariate Haseman-Elston test (NMHE) (Elston et al. 2000). We show that the power of NMHE was, for different simulation setups, identical or higher than that of MHE. In the bivariate case, the power of the NMHE method was somewhat less than that of the computationally intensive maximum likelihood variance components method (Amos et al. 2001). We present comparisons of the empirical distributions of the NMHE test to its limiting distributions for a range of numbers of traits. The distribution of the NMHE test appeared to conform satisfactorily to its limiting asymptotic distribution in large samples. Otherwise, empirical critical values for NMHE are somewhat higher than predicted, i.e. the test proposed by Elston et al. (2000) is non-conservative. The use of empirical critical values is therefore recommended for limited sample sizes (less than several hundred families). We also present the results of a linkage analysis performed by the NMHE method on a set of 4 body size-related traits. The method identified meaningful combinations of traits that showed significant linkage on chromosome 2 and suggestive linkage to regions on chromosomes 16 and 17.

Body Constitution↗

Influence of the beta2-adrenergic receptor Arg16Gly polymorphism on longitudinal changes in obesity from childhood through young adulthood in a biracial cohort: the Bogalusa Heart Study.

BACKGROUND: The beta2-adrenergic receptor (ADRB2) plays a major role in regulating energy expenditure by stimulating lipid metabolism in human adipose tissue. Polymorphisms in the ADRB2 gene have been associated with obesity and various weight-related traits in cross-sectional studies of adults, but little is known about the effects of the ADRB2 gene on childhood obesity or the propensity to gain weight over time. OBJECTIVE: To assess the effects of a polymorphism in codon 16 (Arg16-->Gly) of the ADRB2 gene, which has been associated with a decrease in beta2-receptor density and efficiency, on longitudinal changes in obesity from childhood to young adulthood in a biracial cohort. DESIGN: Seven cross-sectional screenings of children and five cross-sectional screenings of young adults who were previously examined as children produced longitudinal data from childhood to young adulthood. METHODS: Height, weight and subscapular and triceps skinfolds were measured by trained examiners following identical protocols over the course of the study. Gender- and age-stratified analyses using random coefficients models were used to examine longitudinal genetic effects on obesity in 1151 African-American and Caucasian males and females who attended an average of six examinations over a 24 y period from childhood to young adulthood. RESULTS: Age-stratified analyses showed no clear genetic relationships with changes in obesity measures over time in females, but an age-dependent association was observed in males, where the relationship between the Arg16Gly polymorphism and obesity became stronger with age. In males who were 4-9 y of age at the beginning of the study in 1973, body mass index (BMI) was 4% higher in Gly/Gly and Arg/Gly males compared to those with Arg/Arg by 26 y of age. Subscapular skinfold measurements in Gly/Gly males became significantly different from Arg/Arg males (20% higher) by age 20. In the oldest male cohort (10-14 y of age in 1973), BMI increased at a significantly greater rate (0.4%/y) in males carrying the Gly16 form of the receptor relative to Arg/Arg males. BMI was significantly different between homozygous genotypes by approximately 26 y of age, and reached 8% higher in Gly/Gly males by age 32. Subscapular skinfolds also increased at a significantly greater rate (2%/y) in Gly/Gly males compared to Arg/Arg males, becoming significantly different (27%) by approximately 22 y of age and reaching a maximum difference of 50% by age 32. CONCLUSIONS: Our data suggest that the beta2-adrenergic receptor is associated with the propensity to gain weight from childhood to young adulthood in males. An increased understanding of genetic influences on the development of obesity may improve the effectiveness of interventions designed to reduce excess body weight and help define the role of genetic factors in diabetes and cardiovascular disease.

Alleles↗

Pharmacogenetic approaches to hypertension therapy: design and rationale for the Genetics of Hypertension Associated Treatment (GenHAT) study.

The Genetics of Hypertension Associated Treatment (GenHAT) study will determine whether variants in hypertension susceptibility genes interact with antihypertensive medication to modify coronary heart disease (CHD) risk in hypertensives. GenHAT is an ancillary study of the Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial, ALLHAT, a double-blind, randomized trial of 42 418 hypertensives, 55 years of age or older, with systolic or diastolic hypertension and one or more risk factors for cardiovascular disease. About 50% are non-white, and about half are female. ALLHAT completes follow-up in March 2002. GenHAT is typing variants in hypertension genes; completion of genotyping is scheduled for 2003. Analysis of gene-treatment interactions in relation to outcomes include CHD, stroke, heart failure, and blood pressure lowering. To our knowledge, GenHAT is the largest pharmacogenetic study ever conducted. An added strength is its ability to link gene-treatment interactions with important clinical outcomes across diverse ethnic and gender groups.

Humans↗

Familial thoracic aortic aneurysms and dissections: genetic heterogeneity with a major locus mapping to 5q13-14.

BACKGROUND: Aneurysms and dissections affecting the ascending aorta are associated primarily with degeneration of the aortic media, called medial necrosis. Families identified with dominant inheritance of thoracic aortic aneurysms and dissections (TAA/dissections) indicate that single gene mutations can cause medial necrosis in the absence of an associated syndrome. METHODS AND RESULTS: Fifteen families were identified with multiple members with TAAs/dissections. DNA from affected members from 2 of the families was used for a genome-wide search for the location of the defective gene by use of random polymorphic markers. The data were analyzed by the affected-pedigree-member method of linkage analysis. This analysis revealed 3 chromosomal loci with multiple markers demonstrating evidence of linkage to the phenotype. Linkage analysis using further markers in these regions and DNA from 15 families confirmed linkage of some of the families to 5q13-14. Genetic heterogeneity for the condition was confirmed by a heterogeneity test. Data from 9 families with the highest conditional probability of being linked to 5q were used to calculate the pairwise and multipoint logarithm of the odds (LOD) scores, with a maximum LOD of 4.74, with no recombination being obtained for the marker D5S2029. In 6 families, the phenotype was not linked to the 5q locus. CONCLUSIONS: A major locus for familial TAAs and dissections maps to 5q13-14, with the majority (9 of 15) of the families identified demonstrating evidence of linkage to this locus. The condition is genetically heterogeneous, with 6 families not demonstrating evidence of linkage to any loci previously associated with aneurysm formation.

Aortic Dissection↗

Thrombomodulin Ala455Val polymorphism and risk of coronary heart disease.

BACKGROUND: Thrombomodulin (TM) is expressed on the endothelial surface and plays an important role in vasoprotection. A common polymorphism of TM at amino acid position 455 with an alanine (A) to valine (V) transition was previously reported to be associated cross-sectionally with acute myocardial infarction. Whether this single nucleotide polymorphism predicts risk of developing coronary heart disease (CHD) is unclear. METHODS AND RESULTS: Within a large cohort study, we identified 467 incident CHD cases during an average of 5 years of follow-up. We determined TM-455 genotypes on 376 CHD cases (23% black, 77% white) and a reference sample of 461. The AA genotype was significantly more prevalent in noncases than in cases (P:=0.016). The prevalences of the AA genotype in noncase blacks and whites were 93% and 67%, respectively. The AA genotype frequency was significantly reduced in black cases versus noncases (P:=0.018). It was also lower in white cases than in noncases, but the difference was not statistically significant (P:=0.066). Weighted proportional hazards regression analysis after adjustment for age, sex, and other CHD risk factors showed that having the V allele increased risk of CHD by 6.1-fold (risk ratio 6.1, 95% CI 1.7 to 22.9) in blacks but did not significantly increase the risk in whites. CONCLUSIONS: The TM A455V polymorphism predicts risk of developing CHD in blacks.

Adult↗

Linkage disequilibrium structure and its impact on the localization of a candidate functional mutation.

We have used the unblinded MG1/Q1 Genetic Analysis Workshop 12 simulated data as a model system for investigating the use of linkage disequilibrium structure and simple genotype-phenotype associations to identify candidate functional mutations within a gene of interest. Analysis of the pattern of pairwise linkage disequilibrium indicated three groups of single-nucleotide polymorphisms for which the linkage disequilibrium was high between sites within a group, but lower between sites of different groups. Using linear regression to predict levels of the trait Q1 showed that the known functional site, 5782, was usually not the best genetic predictor of Q1, but sites belonging to the same group as 5782 (i.e., group 2) were always included in the prediction model. In 49 out of the 50 replicates, the functional site was not the best predictor of the trait. Finally, more detailed analyses demonstrate that the relationship between the adjusted R2 for the marker in the prediction model and its disequilibrium with 5782 was linear with the intercept at the origin and terminating at the R2 value for the known functional mutation when the disequilibrium is maximal. These data indicate that simple association studies will not identify the functional mutation, but rather will identify candidate functional mutations that are in very tight linkage disequilibrium with the functional mutation.

Chromosome Mapping↗

High-throughput multiplex SNP genotyping with MALDI-TOF mass spectrometry: practice, problems and promise.

Single nucleotide polymorphisms (SNPs) are currently being identified and mapped at a remarkable pace, providing a rich genetic resource with vast potential for disease gene discovery, pharmacogenetics, and understanding the origins of modern humans. High-throughput, cost effective genotyping methods are essential in order to make the most advantageous and immediate use of these SNP data. We have incorporated the use of matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF) in our laboratory as a tool for differentiating genotypes based on the mass of the variant DNA sequence, and have utilized this method for production scale SNP genotyping. We have combined a 4 microl PCR amplification reaction using 3 ng of genomic DNA with a secondary enzymatic reaction (mini-sequencing) containing oligonucleotide primers that anneal immediately upstream of the polymorphic site, dideoxynucleotides, and a thermostable polymerase used to extend the PCR product by a single base pair. Mass spectrometry (MS) analysis of mini-sequencing reactions was performed using a MALDI-TOF instrument (Voyager-DE, Perseptive Biosystems, Framingham, MA). We performed both single and multiplex PCR and mini-sequencing reactions, and genotyped seven different variant sites in a random sample of 989 individuals. Genotypes generated with MS methods were compared with genotypes produced using a 5' exonuclease fluorescence-based assay (Taqman, Applied Biosystems, Foster City, CA) and a gel-based genotyping protocol. Because multiple polymorphisms can be detected in a single reaction, the MS technique provides a cost-effective and efficient method for high-throughput genotyping.

Alleles↗

Feature (gene) selection in gene expression-based tumor classification.

There is increasing interest in changing the emphasis of tumor classification from morphologic to molecular. Gene expression profiles may offer more information than morphology and provide an alternative to morphology-based tumor classification systems. Gene selection involves a search for gene subsets that are able to discriminate tumor tissue from normal tissue, and may have either clear biological interpretation or some implication in the molecular mechanism of the tumorigenesis. Gene selection is a fundamental issue in gene expression-based tumor classification. In the formation of a discriminant rule, the number of genes is large relative to the number of tissue samples. Too many genes can harm the performance of the tumor classification system and increase the cost as well. In this report, we discuss criteria and illustrate techniques for reducing the number of genes and selecting an optimal (or near optimal) subset of genes from an initial set of genes for tumor classification. The practical advantages of gene selection over other methods of reducing the dimensionality (e.g., principal components), include its simplicity, future cost savings, and higher likelihood of being adopted in a clinical setting. We analyze the expression profiles of 2000 genes in 22 normal and 40 colon tumor tissues, 5776 sequences in 14 human mammary epithelial cells and 13 breast tumors, and 6817 genes in 47 acute lymphoblastic leukemia and 25 acute myeloid leukemia samples. Through these three examples, we show that using 2 or 3 genes can achieve more than 90% accuracy of classification. This result implies that after initial investigation of tumor classification using microarrays, a small number of selected genes may be used as biomarkers for tumor classification, or may have some relevance in tumor development and serve as a potential drug target. In this report we also show that stepwise Fisher's linear discriminant function is a practicable method for gene expression-based tumor classification.

Humans↗