- Research article
- Open Access
Genome-wide expression quantitative trait loci (eQTL) analysis in maize
© Holloway et al; licensee BioMed Central Ltd. 2011
Received: 11 March 2011
Accepted: 30 June 2011
Published: 30 June 2011
Expression QTL analyses have shed light on transcriptional regulation in numerous species of plants, animals, and yeasts. These microarray-based analyses identify regulators of gene expression as either cis-acting factors that regulate proximal genes, or trans-acting factors that function through a variety of mechanisms to affect transcript abundance of unlinked genes.
A hydroponics-based genetical genomics study in roots of a Zea mays IBM2 Syn10 double haploid population identified tens of thousands of cis-acting and trans-acting eQTL. Cases of false-positive eQTL, which results from the lack of complete genomic sequences from both parental genomes, were described. A candidate gene for a trans-acting regulatory factor was identified through positional cloning. The unexpected regulatory function of a class I glutamine amidotransferase controls the expression of an ABA 8'-hydroxylase pseudogene.
Identification of a candidate gene underlying a trans-eQTL demonstrated the feasibility of eQTL cloning in maize and could help to understand the mechanism of gene expression regulation. Lack of complete genome sequences from both parents could cause the identification of false-positive cis- and trans-acting eQTL.
Genomic sequencing of crop species has shed light on causative relationships between sequence polymorphisms and traits of agronomic interest. Ongoing efforts in maize QTL (quantitative trait locus/loci) mapping have identified genetic intervals whose underlying genes variably contribute to interesting phenotypes such as oil content , root architecture , and pest resistance . While many trait variations (quantitative and qualitative) result from amino acid differences [1, 4], gene expression differences can also result in observable phenotypes . Considering the burgeoning fields of epigenetics and transcriptomics, analysis of gene expression regulation is playing an important role in understanding gene interactions that lead to traits of interest.
The concept of "genetical genomics"  was proposed with the advance of high throughput gene expression profiling technologies. In traditional QTL analyses, linkage mapping leads to the detection of genomic regions which are associated with phenotypic variations within a population. Genetical genomics employs this same approach, except that the phenotypes are levels in gene expression resulting in the detection of expression QTL (eQTL). eQTL do not necessarily result from sequence polymorphisms proximal to the gene being measured (cis-acting) but could result from differences in genes unlinked to the target. In these cases, the eQTL function in a trans-acting manner.
The field of genetical genomics has allowed eQTL analysis within mapping populations in a multitude of species of plants [7–10], yeast [11, 12], and mammals [13–16]. An Arabidopsis study  describes that while there are more trans-acting factors within the genome, cis-acting elements are more significant and therefore stronger in regulatory ability than those acting in trans. The results suggest a generalization that while multiple trans-acting factors can each weakly contribute to the total expression regulation of a given gene, a single cis-acting variation plays a far greater role.
The mapping and positional cloning of a trans-acting eQTL may reveal an expression regulator such as a transcription factor or small regulatory RNA. While many eQTL have been identified, few trans-acting factors have been cloned. Yvert et al.  mapped and cloned two yeast trans-eQTL that regulate genes known to be involved in pathways regulating pheromone response and daughter cell separation following budding. Interestingly, neither of the causative genes functions as expression regulators such as transcription factors nor through other expected mechanisms. However, their work suggests that the continued high-throughput eQTL studies will identify novel genes to better understand the regulation of known biochemical pathways. Master regulators such as LEAFY, an Arabidopsis regulator of at least seven components involved in reproductive development , were previously only identifiable through mutagenesis. Genome-wide trans-eQTL analyses may identify more master regulators that function to control several components of a single pathway.
We report a genome wide eQTL analysis in a highly utilized maize mapping population. We successfully identify both cis-acting and trans-acting genetic elements that cooperate to regulate gene expression in maize crown roots, and describe the pitfalls of detecting false cis- or trans-acting eQTL in the absence of perfect genomic sequences from both parents. In addition to this genome-wide analysis of regulating factors, we have positionally cloned a trans-acting factor.
Global analysis of cis-acting and trans-acting eQTL in maize root
Quantity and mode of action of significant (≤10-6) eQTL peaks mapping in biological replicate 1
KS p-value 10 -6 - 10 -15
KS p-value 10 -15 - 10 -30
High reproducibility of biological replicates
Cooperative regulation of gene expression
Cooperative expression regulation of 175 mapped genes
secondary cis-eQTL per primary eQTL
secondary trans-eQTL per primary eQTL
tertiary trans-eQTL per primary eQTL
Total verifiable peaks identified
Lack of perfect genome sequences from both parental lines caused the identification of false cis- and trans-acting eQTL
It is obvious that some of the cis- and trans-acting eQTL identified are false due to the unavailability of complete genome sequences for both parental lines. It is not clear how extensive the problem is, although it seems the strongest eQTL are more prone to be false positive than the weaker ones.
Differential gene expression confirms presence/absence variations
In addition to nucleotide polymorphisms or minor insertion/deletion differences between the B73 and Mo17 parent genomes, several studies have identified large deletions in the Mo17 genome as compared to B73 [23, 24]. Of particular interest is a deletion spanning many BACs on the short arm of chromosome 6 which contains at least 23 genes and pseudogenes . eQTL mapping using the IBM2 Syn10 population found 28 strong (KS p-value range = 10-21 - 10-10) cis-acting eQTL, measuring 28 probes (from 26 annotated genes) located on the BACs in question. For the genes in the region, B73 in general shows robust expression while Mo17 expression is essentially off (data not shown). The eQTL results are consistent with the presence/absence variations (PAV) detected through array genome hybridization.
Identification of a trans-acting regulator by map-based cloning
Glutamine amidotransferase genes are well characterized players in several biosynthesis pathways including the purine, pyrimidine, histidine, tryptophan, and arginine pathways , however, a transcript-abundance mediating function has never been published. While steady state transcript abundance is determined by many factors including transcription factors, enhancers, mRNA degradation regulators, and rate of translation , the role of glutamine amidotransferase in ABA 8'-hydroxylase transcript abundance remains elusive. Despite the role that glutamine amidotransferase plays on ABA 8'-hydroxylase transcription, no other eQTL mapped to the same genetic region in this population suggesting that any broad transcriptional regulatory function is unlikely.
eQTL are determined by expression phenotypes, and not based on physiological or morphological phenotypes as in classic QTL identification. Therefore, further analysis of ABA biosynthesis in the IBM2 Syn10 population could prove a physiological function downstream of transcript abundance in regards to its trans-regulation. In anticipation of ABA biosynthesis analysis, the ABA 8'-hydroxylase gene being regulated by glutamine amidotransferase was sequenced from B73 and Mo17 cDNA. Interestingly, the results showed that the annotated ABA 8'-hydroxylase gene was actually a product of genomic shuffling that occurred sometime prior to the genetic diversification of B73 and Mo17. In both parentages, the ABA 8'-hydroxylase pseudogene (ABA-8' p ) is a chimera of the functional maize ABA 8'-hydroxylase gene fragment, non-genic genomic sequence originating from chromosome 10, and repetitive genomic sequence (Figure 7B). Thus it seems that a pseudogene expression level is regulated. The ABA 8' p is not present in all maize inbred lines. Actually, validation of glutamine amidotransferase as the candidate gene for the eQTL was hampered by the absence of this pseudogene in certain lines (data not shown). Whether this pseudogene or its expression regulation has any physiological function remains to be determined.
While the field of genomics has afforded scientists with access to genomic sequences of countless numbers of model species, strains, and lines, our understanding of the function of those sequences remains rather limited. Forward and reverse genetics have given meaning to sequence polymorphisms in a fair number of genes, but in addition to gene function analysis, it is equally important to understand how and when those genes are activated and the roles the translated proteins play within biochemical pathways. Only through the continued efforts in the fields of transcriptomics and proteomics can the full power of genomics be realized. Expression QTL studies are generating vast amounts of data from the perspective of gene regulation, both from cis-acting elements and trans-acting factors, that begin to fill in gaps in the understanding of transcriptional regulation and gene interactions. Results from our study help to elucidate the genome-wide expression regulation in play during the development of crown root tips in maize. Despite stringent statistical analysis, we identified more than 10,000 eQTL that function through both cis- and trans-acting mechanisms. In addition to the identification of cis- and trans-factors, we described the relative regulatory contribution each of those factors plays by means of a KS p-value. Despite our expectations, as well as those set forth in previously published maize eQTL studies [28, 29], we identified vastly more cis-acting eQTL than trans-suggesting that the most significant eQTL will act in cis for most genes. The statistical methodology and stringency we employed was designed to minimize false positive eQTL from the analysis, however at the probable expense of the lesser significant eQTL which will most likely function in trans . Additionally, the definition of trans can be "arbitrarily" set for each study. While we defined trans-acting to mean any regulatory element greater than 10 cM from the target, others have set the boundary at 5 cM . Either limit is appropriate for maize eQTL studies but will affect the classification and quantification of trans-eQTL.
Among the tens of thousands of cis- and trans-acting eQTL identified in the current study, we have demonstrated that some of them are false-positives due to the lack of complete genome sequences from both parents of the mapping population. Polymorphism in the microarray probe regions, which could affect hybridization intensities for mRNA quantification, will lead to the occurrence of false cis-acting eQTL. Likewise, sequencing gaps at the trans-acting eQTL regions could result in the detection of false trans-acting eQTL. With these mechanisms, the stronger, large-effect eQTL (with low p-values) are more prone to be false positives than the weaker eQTL. We estimated that 35% of all the cis-eQTL is false positive, but the false trans-eQTL discovery rate is unknown in the absence of the complete genome sequences. It is likely that the whole-genome eQTL analysis reported in other systems could suffer similar false positive issues, and caution should be taken when interpreting the results and selecting eQTL for further analysis.
Although tens of thousands of eQTL have been reported, very few have been cloned and characterized, especially in higher organisms where positional cloning could still be challenging. To test the feasibility of eQTL cloning in maize and to understand the regulatory mechanism of transcript abundance, a strong trans-eQTL for a putative ABA hydroxylase was selected for further mapping and cloning. The eQTL was fine mapped to a very small physical interval (186 bp) and a putative glutamine amidotransferase gene was identified as the candidate gene for this eQTL. Unfortunately, the target gene identified turned out to be a pseudogene. It is not clear if this ABA hydroxylase pseudogene has any physiological function, why the expression of a pseudogene is regulated, or how the two genes interact with each other. Nevertheless, we have clearly demonstrated the feasibility of cloning trans-eQTL with large effect in maize. The cloning of trans-eQTL would help to understand the mechanism of transcript abundance regulation and identify regulatory genes for biochemical pathways.
The IBM2 SYN10 population is a set of 360 doubled haploid lines from a randomly mated population derived from B73 and Mo17 . Having undergone 10 generations of random intermating/recombination, these DH lines exhibited a high degree of phenotypic variability and high frequency of recombination. We were able to fine map an eQTL into a 186 bp interval with only 135 IBM2 SYN10 DH lines. Although recombination frequency varies widely across the genome and more genomic loci need to be analyzed before this population can be better assessed for recombination frequency, the results suggest that the IBM Syn10 population is suitable for high resolution eQTL mapping and recombination studies.
As in previous reports, we have shown that most of the strong eQTL act in cis. Moreover, for the relatively small number of strong trans-eQTL detected, some of them could be false positive. Therefore, it may be a challenge to identify a large number of strong trans-eQTL which are more amenable for positional cloning. A recently study in human B cells which maps the regulatory elements that influence radiation-induced changes in gene expression indicated that nearly all the strong regulators act in trans to influence the expression of their target genes . Therefore, instead of mapping the steady-state level of mRNA under one constant condition, mapping eQTL which regulate the differential responsiveness in gene expression to biotic and abiotic factors could be a promising approach to enrich and identify strong trans-eQTL. These trans-eQTL should be ideal targets for cloning. They are important for understanding plant responses to biotic and abiotic stresses, and technically feasible to isolate.
We have shown the feasibility of eQTL analysis as a means to identify, clone and analyze trans-acting regulatory factors through large-scale screening analysis. A glutamine amidotransferase regulator was identified as a trans-acting factor. Harnessing the regulatory function of trans-acting factors could allow for better control of important agronomical genes. We also described the pitfall of identifying false-positive eQTL in the absence of complete genome sequences, which has broad implications in similar eQTL studies.
Plants were germinated and grown in Turface ®MVP® contained in Deepot™ plastic conical pot and tray systems (Stuewe and Sons, Tangent, OR (D25L [pots], D50T [trays])). The hydroponic system consisted of stainless steel tanks (24"w × 24"d × 15"h) with a bottom-filling pumping and drainage system and overflow drain at the level of the top of the tubes/Turface®MVP®. Seeds were planted approximately 2 cm below the top of the tube/Turface®MVP® and germinated for one week in diH2O submerged so that the seeds were just above the water line. After germination, the water was drained and a media pumping system was initiated. A humidity cover was in place from the initial germination until plants were approximately 6 inches tall. Modified Hoagland's Solution (1 mM KH2PO4, 2.5 mM KNO3, 1.25 mM Ca(NO3)2, 1 mM MgSO4, 0.75 mM CaCl2, SPRINT330 such that 0.006 mM with respect to Iron, 0.03 mM FeSO4, 1 μM H3Bo3, 1 μM MnCl2, 1 μM ZnSo4, 1 μM CuSO4, 1 μM NaMoO4) was pumped into the tanks until the tubes were submerged but the vegetative growth remained dry. The nutrient solution remained in the tank for approximately 3 minutes before draining. The submersion/drainage system repeated every 3 hours for the 3 weeks following germination. Growth chamber conditions were maintained at 50% humidity with a 16 h day (26°C) and an 8 h night (22°C). Light levels ranged from approximately 350 μM/m2/s at the base of the plant to 500 μM/m2/s at the top of the leaves at the time of harvest.
Plant materials and tissue preparation
For each of two biological replicates, tissue was collected from B73 and Mo17 parental controls and the IBM2 Syn10 doubled haploid mapping population  at 4 weeks after germination (V5 stage). The IBM Syn10 population consisted of 360 doubled haploid lines resulting from 10 generations of intermating, followed by a double haploid process creating highly recombinant, yet fixed alleles. One hundred thirty-five IBM2 Syn10 lines were used in this study. The tissue sampled was the 1.5 cm tip of all crown roots that had developed at the time of harvest, representing the most metabolically active region based on whole-tissue visualization of triphenyl tetrazolium chloride staining . Tissues were flash frozen in liquid nitrogen and stored at -80°C. They were manually ground into a fine powder using a mortar and pestle on dry ice.
Total RNA was isolated from frozen ground tissue (SQ Tissue Kit, Omega Biotek) and treated with DNase-I followed by polyA RNA isolation (Illustra mRNA Purification Kit, GE Biosciences) for all samples. The total RNA and polyA RNA samples were visualized and quantified on Agilent's Bioanalyzer to check for degradation and to determine the concentration. Each mRNA sample was made into double stranded DNA, amplified by an in-vitro transcription reaction and labeled with Cy3 or Cy5 fluorescent dyes using Agilent's Low RNA Linear Amp Kit. The cRNA product was purified with Agencourt's RNAClean Kit that utilizes SPRI (Solid Phase Reversible Immobilization) paramagnetic bead-based technology. Overnight hybridizations were performed with equal amounts of labeled cRNA to custom 2 × 105K  Maize Oligo Microarrays from Agilent Technologies (Palo Alto, CA) according to Agilent's Two-Color Microarray-Based Gene Expression Analysis protocol. After hybridization, the microarray slides were washed and immediately scanned with Agilent's G2505B DNA Microarray Scanner at two laser power settings (100% and 10%). The images were visually inspected for image artifacts and feature intensities were extracted, filtered, and normalized with Agilent's Feature Extraction Software (v 9.5.1). Further quality control analysis was performed using data analysis tools in Rosetta's Resolver Database. The Agilent microarray expression data (raw and processed) are available from GEO (Series Accession GSE29964).
Statistical analysis and eQTL mapping
Microarray intensity data were determined from each of the two channels using the Rosetta Resolver Split-Ration method [32, 33], were exported from the Rosetta Resolver Database and analyzed using software developed in-house. The following is a brief summary of the concepts that underlie our data analysis methods. A multidimensional, weighted least-squares method was used to obtain normalization parameters for the data based on affine transformations, an effective normalization approach according to measurement theory  and fluorescence instrumentation considerations . A further implication is that various stochastic effects, both instrumental and biological, give rise to an overall intensity dependent noise [36–38]. Data was analyzed using the intensity representation to preserve the noise characteristics and estimate statistical significance, facilitating the use of linear least-square methods for data analysis. The intensity distribution shows that about 20 percent of the data from a microarray typically corresponds to background. Therefore an additive correction was applied to adjust the average background signal for a microarray to zero, prior to normalization. As described above, gene expression was measured using a single color experiment design for two biological replicates of the SX19 Syn10 population. The population was also genotyped at 1731 SNP markers using the Illumina assay. Initially, 50 eQTL were individually mapped using the Knott-Haley regression method for interval mapping (Windows QTL Cartographer, Statistical Genetics and Bioinformatics, North Carolina State University, USA) to confirm subsequent analyses where the Kolmogorov-Smirnov (KS) test was used to test each marker for association with normalized intensity from each gene signature. The map positions of eQTL were assigned based on the most significant KS p-value calculated for each probe by genome wide association scanning, therefore assigning only one eQTL per probe. Various methods have been proposed for controlling the false positive rate on the basis of uniform distributions of p-values [39, 40]. However, the problem of adjusting for bias arising from correlation in hypotheses remains a challenge [41, 42]. Thus, we adopt a heuristic approach based on the analysis of the distribution of KS p-values for all eQTL, which gives a threshold of 10-5 for controlling the false discovery rate. Furthermore, association testing methodology including those used here rely on assumptions that may not hold, particularly with regard to noise. Therefore, a slightly more conservative threshold of 10-6 was used for identifying eQTL, and to obtain a lower false positive rate for the purposes of this study. To test overall reproducibility, data for each biological replicate were analyzed separately for eQTL including eQTL map position, KS p-value, and relative microarray intensity.
Map-based cloning of a trans-acting regulator
A strong trans-acting eQTL was selected for fine mapping and cloning. Genomic DNA from the IMB2 Syn10 population and parental controls were purified from leaf tissue using Gentra® Puregene® (Qiagen, Valencia, CA)  modified in scale of preparation. Fine mapping of the interval was initiated at the genetic position determined by eQTL mapping analysis of the microarray expression data. Total RNA for gene expression validation was purified as above and reverse transcribed into cDNA using QuantiTect® Reverse Transcription Kit (Qiagen, Valencia, CA). The B73 and Mo17 alleles of the candidate gene, glutamine amidotransferase, were amplified from cDNA using the primers 5'-CCTAAGACATCCCAATTTCCTC and 5'-GTCGCCTCCATCTCCATTC, cloned into the pCR®2.1.TOPO® TA vector (Invitrogen, Carlsbad, CA), and were confirmed by sequencing. The eQTL was re-mapped using the IBM2 Syn10 population based on a quantitative RT-PCR phenotype for the target previously measured by the microarray (a pseudogene of ABA-8'Hydroxylase) using the primers 5'-GCGTTGAACACTTGGACCAC-3' and 5'-TGGAAGGTGTTGCCCCTGTT-3'.
We would like to thank Dr. Andre Beló and Kevin Fengler for insightful discussion in regard to genomic analysis, and physical and genetic maps. We would like to thank April Leonard and Nathan Uhlmann for technical assistance in the laboratory and the growth chambers. We thank Brian Zeka for the RNA isolations and John Nau for the array hybridizations. We would like to thank April Leonard for critical commentary on the manuscript. We thank Dr. Scott Tingey for intellectual and scientific support for the project. B.H. was supported by a Discovery Grant from Pioneer Hi-Bred Inc.
- Zheng P, Allen WB, Roesler K, Williams ME, Zhang S, Li J, Glassman K, Ranch J, Nubel D, Solawetz W, Bhattramakki D, Llaca V, Deschamps S, Zhong GY, Tarczynski MC, Shen B: A phenylalanine in DGAT is a key determinant of oil content and composition in maize. Nat Genet. 2008, 40 (3): 367-372. 10.1038/ng.85.View ArticlePubMedGoogle Scholar
- Ruta N, Liedgens M, Fracheboud Y, Stamp P, Hund A: QTLs for the elongation of axile and lateral roots of maize in response to low water potential. Theor Appl Genet. 2009Google Scholar
- Balint-Kurti PJ, Carson ML: Analysis of quantitative trait Loci for resistance to southern leaf blight in juvenile maize. Phytopathology. 2006, 96 (3): 221-225. 10.1094/PHYTO-96-0221.View ArticlePubMedGoogle Scholar
- Wang H, Nussbaum-Wagler T, Li B, Zhao Q, Vigouroux Y, Faller M, Bomblies K, Lukens L, Doebley JF: The origin of the naked grains of maize. Nature. 2005, 436 (7051): 714-719. 10.1038/nature03863.View ArticlePubMedPubMed CentralGoogle Scholar
- Doebley J, Stec A, Hubbard L: The evolution of apical dominance in maize. Nature. 1997, 386 (6624): 485-488. 10.1038/386485a0.View ArticlePubMedGoogle Scholar
- Jansen RC, Nap JP: Genetical genomics: the added value from segregation. Trends Genet. 2001, 17 (7): 388-391. 10.1016/S0168-9525(01)02310-1.View ArticlePubMedGoogle Scholar
- Kliebenstein DJ, West MA, van Leeuwen H, Loudet O, Doerge RW, St Clair DA: Identification of QTLs controlling gene expression networks defined a priori. BMC Bioinformatics. 2006, 7: 308-10.1186/1471-2105-7-308.View ArticlePubMedPubMed CentralGoogle Scholar
- Wise RP, Moscou MJ, Bogdanove AJ, Whitham SA: Transcript profiling in host-pathogen interactions. Annu Rev Phytopathol. 2007, 45: 329-369. 10.1146/annurev.phyto.45.011107.143944.View ArticlePubMedGoogle Scholar
- Schadt EE, Molony C, Chudin E, Hao K, Yang X, Lum PY, Kasarskis A, Zhang B, Wang S, Suver C, Zhu J, Millstein J, Sieberts S, Lamb J, GuhaThakurta D, Derry J, Storey JD, Avila-Campillo I, Kruger MJ, Johnson JM, Rohl CA, van Nas A, Mehrabian M, Drake TA, Lusis AJ, Smith RC, Guengerich FP, Strom SC, Schuetz E, Rushmore TH, et al: Mapping the genetic architecture of gene expression in human liver. PLoS Biol. 2008, 6 (5): e107-10.1371/journal.pbio.0060107.View ArticlePubMedPubMed CentralGoogle Scholar
- West MA, Kim K, Kliebenstein DJ, van Leeuwen H, Michelmore RW, Doerge RW, St Clair DA: Global eQTL mapping reveals the complex genetic architecture of transcript-level variation in Arabidopsis. Genetics. 2007, 175 (3): 1441-1450.View ArticlePubMedPubMed CentralGoogle Scholar
- Brem RB, Kruglyak L: The landscape of genetic complexity across 5,700 gene expression traits in yeast. Proc Natl Acad Sci USA. 2005, 102 (5): 1572-1577. 10.1073/pnas.0408709102.View ArticlePubMedPubMed CentralGoogle Scholar
- Yvert G, Brem RB, Whittle J, Akey JM, Foss E, Smith EN, Mackelprang R, Kruglyak L: Trans-acting regulatory variation in Saccharomyces cerevisiae and the role of transcription factors. Nat Genet. 2003, 35 (1): 57-64.View ArticlePubMedGoogle Scholar
- Enard W, Khaitovich P, Klose J, Zollner S, Heissig F, Giavalisco P, Nieselt-Struwe K, Muchmore E, Varki A, Ravid R, Doxiadis GM, Bontrop RE, Paabo S: Intra- and interspecific variation in primate gene expression patterns. Science. 2002, 296 (5566): 340-343. 10.1126/science.1068996.View ArticlePubMedGoogle Scholar
- Jia Z, Xu S: Mapping quantitative trait loci for expression abundance. Genetics. 2007, 176 (1): 611-623. 10.1534/genetics.106.065599.View ArticlePubMedPubMed CentralGoogle Scholar
- Kwan T, Benovoy D, Dias C, Gurd S, Provencher C, Beaulieu P, Hudson TJ, Sladek R, Majewski J: Genome-wide analysis of transcript isoform variation in humans. Nat Genet. 2008, 40 (2): 225-231. 10.1038/ng.2007.57.View ArticlePubMedGoogle Scholar
- Cheung VG, Spielman RS, Ewens KG, Weber TM, Morley M, Burdick JT: Mapping determinants of human gene expression by regional and genome-wide association. Nature. 2005, 437 (7063): 1365-1369. 10.1038/nature04244.View ArticlePubMedPubMed CentralGoogle Scholar
- DeCook R, Lall S, Nettleton D, Howell SH: Genetic regulation of gene expression during shoot development in Arabidopsis. Genetics. 2006, 172 (2): 1155-1164.View ArticlePubMedPubMed CentralGoogle Scholar
- William DA, Su Y, Smith MR, Lu M, Baldwin DA, Wagner D: Genomic identification of direct target genes of LEAFY. Proc Natl Acad Sci USA. 2004, 101 (6): 1775-1780. 10.1073/pnas.0307842100.View ArticlePubMedPubMed CentralGoogle Scholar
- Hussain T, Tausand P, Graham G, Ho J: Registration of IBM2 SYN10 Doubled Haploid Mapping Population of Maize. Journal of Plant Registrations. 2007, 1 (81):Google Scholar
- Hayes KR, Beatty M, Meng X, Simmons CR, Habben JE, Danilevskaya ON: Maize Global Transcriptomics Reveals Pervasive Leaf Diurnal Rhythms but Rhythms in Developing Ears Are Largely Limited to the Core Oscillator. PLoS ONE. 2010, 5 (9): e12887-10.1371/journal.pone.0012887.View ArticlePubMedPubMed CentralGoogle Scholar
- van Nas A, Ingram-Drake L, Sinsheimer JS, Wang SS, Schadt EE, Drake T, Lusis AJ: Expression quantitative trait loci: replication, tissue- and sex-specificity in mice. Genetics. 2010, 185 (3): 1059-1068. 10.1534/genetics.110.116087.View ArticlePubMedPubMed CentralGoogle Scholar
- Holloway B, Li B: Expression QTLs: applications for crop improvement. Molecular Breeding. 2010, 26 (3): 381-391. 10.1007/s11032-010-9396-2.View ArticleGoogle Scholar
- Belo A, Beatty MK, Hondred D, Fengler KA, Li B, Rafalski A: Allelic genome structural variations in maize detected by array comparative genome hybridization. Theor Appl Genet. 2010, 120 (2): 355-367. 10.1007/s00122-009-1128-9.View ArticlePubMedGoogle Scholar
- Springer NM, Ying K, Fu Y, Ji T, Yeh CT, Jia Y, Wu W, Richmond T, Kitzman J, Rosenbaum H, Iniguez AL, Barbazuk WB, Jeddeloh JA, Nettleton D, Schnable PS: Maize inbreds exhibit high levels of copy number variation (CNV) and presence/absence variation (PAV) in genome content. PLoS Genet. 2009, 5 (11): e1000734-10.1371/journal.pgen.1000734.View ArticlePubMedPubMed CentralGoogle Scholar
- Cone K, McMullen M, Bi IV, Davis GL, Yim YS, Gardiner JM, Polacco MI, Sanchez-Villeda H, Fang Z, Schroeder SG, Havermann SA, Bowers JE, Paterson AH, Soderlund CA, Engler FW, Wing RA, Coe EH: Genetic, physical, and informatics resources for maize. On the road to an integrated map. (0032-0889 (Print))Google Scholar
- Massiere F, Badet-Denisot MA: The mechanism of glutamine-dependent amidotransferases. Cell Mol Life Sci. 1998, 54 (3): 205-222. 10.1007/s000180050145.View ArticlePubMedGoogle Scholar
- Williams RB, Chan EK, Cowley MJ, Little PF: The influence of genetic variation on gene expression. Genome Res. 2007, 17 (12): 1707-1716. 10.1101/gr.6981507.View ArticlePubMedGoogle Scholar
- Shi C, Uzarowska A, Ouzunova M, Landbeck M, Wenzel G, Lubberstedt T: Identification of candidate genes associated with cell wall digestibility and eQTL (expression quantitative trait loci) analysis in a Flint × Flint maize recombinant inbred line population. BMC Genomics. 2007, 8: 22-10.1186/1471-2164-8-22.View ArticlePubMedPubMed CentralGoogle Scholar
- Swanson-Wagner RA, DeCook R, Jia Y, Bancroft T, Ji T, Zhao X, Nettleton D, Schnable PS: Paternal dominance of trans-eQTL influences gene expression patterns in maize hybrids. Science. 2009, 326: 1071-1072. 10.1126/science.1183463.View ArticleGoogle Scholar
- Smirnov DA, Morley M, Shin E, Spielman RS, Cheung VG: Genetic analysis of radiation-induced changes in human gene expression. Nature. 2009, 459: 587-592. 10.1038/nature07940.View ArticlePubMedPubMed CentralGoogle Scholar
- Chen CW, Yang YW, Lur HS, Tsai YG, Chang MC: A novel function of abscisic acid in the regulation of rice (Oryza sativa L.) root growth and development. Plant Cell Physiol. 2006, 47 (1): 1-13.View ArticlePubMedGoogle Scholar
- Kaufmann K, Wellmer F, Muino JM, Ferrier T, Wuest SE, Kumar V, Serrano-Mislata A, Madueno F, Krajewski P, Meyerowitz EM, Angenent GC, Riechmann JL: Orchestration of floral initiation by APETALA1. Science. 2010, 328: 85-89. 10.1126/science.1185244.View ArticlePubMedGoogle Scholar
- te Velde AA, de Kort F, Sterrenburg E, Pronk I, ten Kate FJ, Hommes DW, van Deventer SJ: Comparative analysis of colonic gene expression of three experimental colitis models mimicking inflammatory bowel disease. Inflamm Bowel Dis. 2007, 13: 325-330. 10.1002/ibd.20079.View ArticlePubMedGoogle Scholar
- Bengtsson H, Hossjer O: Methodological study of affine transformations of gene expression data with proposed robust non-parametric multi-dimensional normalization method. BMC Bioinformatics. 2006, 7: 100-10.1186/1471-2105-7-100.View ArticlePubMedPubMed CentralGoogle Scholar
- Skoog DA, Leary JJ: Principles of Instrumental Analysis. 1992, Harcourt Brace College Pubs., New York, 4Google Scholar
- Quackenbush J: Microarray data normalization and transformation. Nat Genet. 2002, 32 (Suppl): 496-501.View ArticlePubMedGoogle Scholar
- Weng L, Dai H, Zhan Y, He Y, Stepaniants SB, Bassett DE: Rosetta error model for gene expression analysis. Bioinformatics. 2006, 22 (9): 1111-1121. 10.1093/bioinformatics/btl045.View ArticlePubMedGoogle Scholar
- Zeisel A, Amir A, Kostler WJ, Domany E: Intensity dependent estimation of noise in microarrays improves detection of differentially expressed genes. BMC Bioinformatics. 2010, 11: 400-10.1186/1471-2105-11-400.View ArticlePubMedPubMed CentralGoogle Scholar
- Storey JD, Tibshirani R: Statistical significance for genomewide studies. Proc Natl Acad Sci USA. 2003, 100 (16): 9440-9445. 10.1073/pnas.1530509100.View ArticlePubMedPubMed CentralGoogle Scholar
- Wasserman L: All of statistics: a concise course in statistical inference. 2004, New York: SpringerView ArticleGoogle Scholar
- Sarkar S: False Discover and False Nondiscovery Rates in Single-Step Multiple Testing Procedures. The Annals of Statistics. 2006, 34 (1): 394-415. 10.1214/009053605000000778.View ArticleGoogle Scholar
- Schwartzman A, Lin X: The effect of correlation in false discovery rate estimation. biometrika. 2011, 98 (1): 199-214. 10.1093/biomet/asq075.View ArticlePubMedPubMed CentralGoogle Scholar
- Djukanovic V, Orczyk W, Gao H, Sun X, Garrett N, Zhen S, Gordon-Kamm W, Barton J, Lyznik LA: Gene conversion in transgenic maize plants expressing FLP/FRT and Cre/loxP site-specific recombination systems. Plant Biotechnol J. 2006, 4 (3): 345-357. 10.1111/j.1467-7652.2006.00186.x.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.