- Research article
- Open Access
Nonlinear regulation enhances the phenotypic expression of trans- acting genetic polymorphisms
© Gjuvsland et al; licensee BioMed Central Ltd. 2007
- Received: 14 March 2007
- Accepted: 25 July 2007
- Published: 25 July 2007
Genetic variation explains a considerable part of observed phenotypic variation in gene expression networks. This variation has been shown to be located both locally (cis) and distally (trans) to the genes being measured. Here we explore to which degree the phenotypic manifestation of local and distant polymorphisms is a dynamic feature of regulatory design.
By combining mathematical models of gene expression networks with genetic maps and linkage analysis we find that very different network structures and regulatory motifs give similar cis/trans linkage patterns. However, when the shape of the cis- regulatory input functions is more nonlinear or threshold-like, we observe for all networks a dramatic increase in the phenotypic expression of distant compared to local polymorphisms under otherwise equal conditions.
Our findings indicate that genetic variation affecting the form of cis-regulatory input functions may reshape the genotype-phenotype map by changing the relative importance of cis and trans variation. Our approach combining nonlinear dynamic models with statistical genetics opens up for a systematic investigation of how functional genetic variation is translated into phenotypic variation under various systemic conditions.
- Gene Regulatory Network
- Hill Coefficient
- Linkage Pattern
- Gene Expression Network
- Maximal Production Rate
The key disciplinary goal of genetics the last 100 years has been to understand the relationship between genetic variation and phenotypic variation. A series of concepts have been conceived to describe various aspects of the genotype-phenotype map. Many of them reflect the fact that the phenotypic signatures of alleles and genes depend on other alleles and genes (as for example "dominance"  and "epistasis" ). However, these concepts have to be regarded as descriptory rather than explanatory. An explanatory theory capable of linking genetic variation with phenotypic variation in even simple mechanistic terms has yet to emerge . However, there exist a few well studied model systems such as the lambda switch [4, 5] where this link has been described very well.
An empirically sound starting point for such a theory development will be the mRNA phenotype. The genotype-phenotype gap is in this case narrow compared to higher level phenotypes and relatively simple dynamic models can be used to describe much of the systemic behaviour [6–9]. Also, numerous studies have established that a significant fraction of observed inter-individual variability in gene expression is due to cis-linked and trans-linked genetic polymorphisms (reviewed by  and ). How biological systems translate genetic variation into phenotypic variation has recently received some attention [12–16], but there is still an almost completely unrevealed relationship between regulatory polymorphisms, network design principles, and descriptory concepts like cis/trans- linkage, dominance, epistasis and penetrance even at the expression level.
Gjuvsland et al.  showed that gene regulatory networks generate significant amounts of statistical epistasis which depends on the type of feedback regulation involved. Here we address how single gene descriptors and their dependence on (genetically controlled) regulatory design features contribute to the functional epistasis characteristics of mathematical genotype-phenotype maps. Functional epistasis is here used as a common term for describing situations where the phenotypic effect of a genetic substitution (on one or multiple loci) depends on the genetic background, i.e. on the state of other loci in the genotype .
The basic strategy underlying our analysis was to (i) position a fixed number of genes on a genetic map; (ii) introduce dynamic network models for the expression of these genes; (iii) define alleles by a set of model parameters and the equilibrium concentrations of the gene products (with noise added) as these genes' expression phenotype; (iv) introduce genetic variation in the model parameters; (v) make mapping populations of individuals having their expression phenotypes determined by the dynamic network models; and (vi) analyse the populations with the machinery of statistical genetics. This approach opens for a systematic investigation of the phenotypic manifestations of genetic variation as a function of gene network design.
As the steady state abundance of mRNA is dependent on the balance between synthesis and decay, our models include one term for synthesis and one for decay of mRNA.
A polymorphism that has an effect on expression level of a given gene x must transmit this effect through the production and/or degradation term describing the time rate of change of expression of x. This low-resolution modelling approach catches the most important aggregate features of more detailed first-principle models of transcription based on statistical mechanics [19–21] Moreover, the current resolution of empirical data on the existence of non-coding polymorphisms affecting maximal production rates [22–25] and decay rates [26, 27] does not invite to make use of more detailed models of the processes underlying these observations. Thus, by letting the parameters defining production rate and decay rate mediate genetic variation in our genotype-phenotype models we account for a whole range of different, and possibly still unrevealed, mechanistic processes responsible for this variation.
We particularly investigated whether or not different gene network structures create different cis- and trans- linkage patterns, and how the manifestation of phenotypic effects is influenced by the actual form of the cis-regulatory input function . This function (also called gene regulation function ) describes the relationship between the production rates of a given gene product and the concentrations of the regulatory agents controlling these rates. Our motivation for focusing on this function is thus that it is both a basic regulatory design common to all network structure and the prime mediator of trans-acting effects, in both downstream and feedback regulatory relationships . We chose to work with two distinct functional shapes (or modes), one describing ordinary hyperbolic saturation kinetics (being close to linear over much of the concentration span), and one describing moderately nonlinear (sigmoidal) saturation kinetics (Figure 1B). There is solid empirical [25, 34–36] as well as theoretical [19–21, 37–39] support for frequent presence of both modes in eukaryotes, and experimental studies have shown that it is relatively easy in mutational terms to move between a hyperbolic mode and a sigmoidal one [25, 34]. We found that the shape of the cis-regulatory input function has a dramatic influence on the genotype-phenotype map concerning the phenotypic expression of distant compared to local polymorphisms under otherwise equal conditions.
With regard to dominance values, all models contain cases where these are correlated with μ or θ or both (Figure 4). However, the relationship between model parameters and dominance values is dramatically weaker than for additive values. This implies that cis-regulatory variation at a given locus does not relate in a straightforward manner to dominance values associated with its own gene product or on gene products of down-stream loci. Predictors of the dominance variation can be constructed in deterministic models like the ones made use of here, but these predictors will have to include much more extensive information about the system. The variation in dominance effects is thus a more pronounced systemic feature than the variation in additive effects, which in turn implies that the classical definition of dominance as an intralocus interaction [40, 44] should be used with care.
Although cis-regulatory variation is more difficult to detect and understand [45, 46], we have focused on the phenotypic signatures of cis-regulatory variation in transcriptional networks and not taken into account the effect of coding polymorphisms. The rationale for this is that genetical genomics studies in yeast  and mice  strongly suggest that this cis- variation is a very important cause for self-linkage.
Our results apply to a much wider range of regulatory settings than what appears from a superficial inspection of the differential equations (see Methods). This is because a regulatory relationship can be mediated by numerous other gene products influencing a variety of intra- and intercellular processes, a simple example being a transcriptional cascade . As long as all these gene products simply transfer the signal between genes A and B in the form of a well-defined dose-response functional relationship, the complexity of this transduction does not influence our predictions. Sigmoidal gene regulation functions are widely used in models of gene networks. Here we employ the frequently [6, 9, 14, 49] used Hill function (see Methods). Properties of the transcriptional machinery such as multiple transcription factor binding sites, synergy and cooperativity , and fractal kinetics  will contribute to sigmoidal gene regulation functions. Mathematical description of transcription regulation by use of statistical mechanics methodology [19–21] as well as experimental data [25, 35, 36] also suggest that the Hill function is very well suited for describing a whole range of mechanistic processes causing nonlinear transcription responses.
Simulations with genotype-phenotype maps defined by genotypic values is a widely used tool in quantitative genetics, and the main purposes are demonstrating and testing methods for mapping of QTLs [50–52]. Such simulations are very useful for showing differences between various mapping methods, and for identifying weaknesses of current methodologies. The main difference to the approach presented here is that we start out from a dynamic system of genes rather than statistical effects. The genotypic values, which are explicitly defined in the genetic model approach, instead become emergent properties of a biologically interpretable dynamic system. This opens up for a much deeper understanding of functional epistasis aspects  of the genotype-phenotype map in terms of biological processes and mechanisms. This is illustrated by our identification of the cis- regulatory input function as an important provider of functional epistasis to the genotype-phenotype map, which is clearly beyond reach for the standard genetic model approach.
In the case of sigmoidal gene regulation functions, our results (Figure 4) indicate that polymorphisms affecting μ (the ratio between maximal production rate and relative decay rate) will not be directly translated into a QTL effect on the steady state. This opens a new opportunity window for genetical genomics studies. Although frequently considered to be the phenotype level closest to DNA sequence variation , transcript abundance does actually reflect the balance between production and decay rates. These two rates are thus more directly tied up with DNA sequence variation than transcript abundance. Genome-wide studies of decay rates have already been performed in yeast  and human cells , and in principle such data could be used to map rateQTLs in the same way as they are used for expressionQTLs. As our results illustrate how variation can be visible at one phenotypic level and hidden at the next level for systemic reasons, comparing QTLs for rates and expression levels can thus probably be exploited to reveal to which degree systemic silencing of mutations in transcriptional networks is a generic feature or not.
When mathematical models capable of bridging the genotype-phenotype gap are embedded in a framework accounting for the number of individuals, mating structures, allele frequencies, genome-wide variations in recombination frequencies and linkage disequilibrium structures, we possess a tool to understand how various polymorphisms affect phenotypic variation in a population. With our simple models we have here only sketched the potential of this approach, but the methodis likely to be applicable also in more complex settings. Although we in this paper focus on expression networks, there is in principle no limit to how many systemic levels one can include, and how sophisticated the mathematical phenotypes can be . Our approach thus opens up for a systematic investigation of the systemic conditions under which different types of functional genetic variation make detectable contributions to the phenotypic variation of traits of interest to biomedicine, production biology and evolutionary biology. The main constraint will be our capacity to make biologically realistic mathematical descriptions of complex phenotypes over a broad range, not the structural complexities of the genetic variation involved.
Gene regulatory model equations
where θ gives the amount of regulator needed to get 50% of maximal production rate while p determines the steepness of the response. The Hill equation describes Michaelis-Menten regulation for p = 1 and switchlike response as p increases (Figure 1B). We varied p between simulations, but within replicates of a particular scenario p is fixed both between alleles and across regulatory actions. If the regulatory effect is inhibitory, the regulatory function 1 - H(y,θ,p) is used.
Six diploid mathematical models of the interaction diagrams in Figure 1A were made using the sigmoid formalism. In all the equations j = 1,2 and y j = xj 1+ xj 2, i = 1,2,3.
The same genetic map was used for all simulations. This map contained five 100 cM chromosomes, and marker loci were spaced equidistant at each 5 cM along the chromosomes. The three genes were placed at the three first chromosomes, gene 1 at c1-42.5 cM, gene 2 at c2-22.5 cM and gene 3 at c3-57.5 cM. Haldane's mapping function was used to compute recombination rates between loci.
For each of the six gene regulatory network models two simulations were run with different Hill coefficients in the regulatory functions. Hill coefficients 1 and 5 were used for models 2–6, but for model 1 Hill coefficient 5 gave cyclic behaviour, and the steepness was reduced to Hill coefficient 2. We started by sampling allelic parameter values from independent uniform distribution of all three types of heritable model parameters, maximal production rate α, threshold for regulation θ, and relative decay rate γ, such that 70 ≤ α ≤ 150, 5 ≤ θ ≤ 15, and 10 ≤ γ ≤ 15. To allele i of gene j we associated α ij , γ ij and one or two θ kij depending on the model and gene. For a given diploid genotype consisting of parameter values for two alleles at each of the three genes the resulting system of equations was solved to find the stable equilibrium values for all three genes, and these simulated expression levels were used as the genotype's contribution to the phenotype. To get the individual phenotype record, independent normally distributed noise with mean 0 and variance 25 was added to the genotypic contribution. For each network model and Hill coefficient we created a set of mapping populations by sampling allelic parameter values for 40 fully homozygous lines, half (P1-lines) of these lines were homozygous 11 at all marker loci, the other half (P2-lines) were homozygous 22. Finally, each P1-line was crossed to each P2-line in an F2-cross. For each gene regulatory model and Hill coefficient this gave 400 F2 populations for the genetic analysis.
Haley-Knott regression was done using the function scanone in the R\qtl package , while Multitrait IM was done with the function JZmapqtl in QTL Cartographer [60, 61]. Genome wide 5% significance thresholds for LOD and LR scores were set by applying both methods to 2500 F2 populations of 250 individuals, using the same genetic map as in the simulations, but with only environmental noise contributing to the expression levels. For both methods the test statistic was computed every 2.5 cM along the whole genetic map. If a test statistic exceeded the threshold a QTL was inferred, however, at most one QTL was flagged from each chromosome.
Evaluation of QTL results
By comparing the positions of the three genes underlying the simulated phenotypes to flagged QTL positions, genes were divided into two groups: detected and not detected. A gene was classified as detected if a significant QTL was flagged on the chromosome at which the gene resided and the QTL peak was in the same marker bracket as the gene or in one of the neighbour brackets (i.e. ≤ 7.5 cM away from the gene), otherwise the gene was classified as not detected.
For each model and Hill coefficient, vectors of genotypic values and allelic differences in values of network model parameters were collected. These vectors had 400 elements, each representing one F2 population. Correlation coefficients for all pair-wise combinations of statistical and network model parameters were computed. The significance of each of these correlation coefficients was evaluated by computing the correlation coefficients of 1000 permutations, reshuffling elements at random within one of the vectors. Correlation coefficients falling inside the interval observed in the permuted datasets were set to zero.
This study was supported by the National Programme for Research in Functional Genomics in Norway (FUGE) in the Research Council of Norway (grant no. NFR153302). We thank two anonymous reviewers for helpful comments on the manuscript.
- Mendel G: Versuche über Phlanzen-Hybriden. Verhandlungen des naturforschenden Vereines in Brünn. 1865, Band IV: 3-47.Google Scholar
- Bateson W: Mendel's Principles of Heredity. 1909, Cambridge: Cambridge Univ PressView ArticleGoogle Scholar
- Omholt SW: From bean-bag genetics to feedback genetics: bridging the gap between regulatory biology and classical genetics. Biology of Dominance. Edited by: Veitia RA. 2006, Georgetown, TX: Landes Bioscience, http://www.landesbioscience.com/books//id/887Google Scholar
- Zhu XM, Yin L, Hood L, Ao P: Robustness, stability and efficiency of phage lambda genetic switch: dynamical structure analysis. Journal of Bioinformatics and Computational Biology. 2004, 2: 785-817. 10.1142/S0219720004000946. 10.1142/S0219720004000946PubMedView ArticleGoogle Scholar
- Zhu XM, Yin L, Hood L, Ao P: Calculating biological behaviors of epigenetic states in the phage λ life cycle. Functional & Integrative Genomics. 2004, 4: 188-195. 10.1007/s10142-003-0095-5View ArticleGoogle Scholar
- Becskei A, Seraphin B, Serrano L: Positive feedback in eukaryotic gene networks: cell differentiation by graded to binary response conversion. The EMBO journal. 2001, 20: 2528-2535. 10.1093/emboj/20.10.2528PubMed CentralPubMedView ArticleGoogle Scholar
- Elowitz MB, Leibler S: A synthetic oscillatory network of transcriptional regulators. Nature. 2000, 403: 335-338. 10.1038/35002125PubMedView ArticleGoogle Scholar
- Gardner TS, Cantor CR, Collins JJ: Construction of a genetic toggle switch in Escherichia coli. Nature. 2000, 403: 339-342. 10.1038/35002131PubMedView ArticleGoogle Scholar
- Rosenfeld N, Elowitz MB, Alon U: Negative autoregulation speeds the response times of transcription networks. Journal of Molecular Biology. 2002, 323: 785-793. 10.1016/S0022-2836(02)00994-4PubMedView ArticleGoogle Scholar
- Stamatoyannopoulos JA: The genomics of gene expression. Genomics. 2004, 84: 449-457. 10.1016/j.ygeno.2004.05.002PubMedView ArticleGoogle Scholar
- Rockman MV, Kruglyak L: Genetics of global gene expression. Nature Reviews Genetics. 2006, 7: 862-872. 10.1038/nrg1964PubMedView ArticleGoogle Scholar
- Frank SA: Population and quantitative genetics of regulatory networks. Journal of Theoretical Biology. 1999, 197: 281-294. 10.1006/jtbi.1998.0872PubMedView ArticleGoogle Scholar
- Moore JH, Williams SM: Traversing the conceptual divide between biological and statistical epistasis: systems biology and a more modern synthesis. BioEssays. 2005, 27: 637-646. 10.1002/bies.20236PubMedView ArticleGoogle Scholar
- Omholt SW, Plahte E, Oyehaug L, Xiang KF: Gene regulatory networks generating the phenomena of additivity, dominance and epistasis. Genetics. 2000, 155: 969-980.PubMed CentralPubMedGoogle Scholar
- Peccoud J, Velden KV, Podlich D, Winkler C, Arthur L, Cooper M: The selective values of alleles in a molecular network model are context dependent. Genetics. 2004, 166: 1715-1725. 10.1534/genetics.166.4.1715PubMed CentralPubMedView ArticleGoogle Scholar
- Welch SM, Dong ZS, Roe JL, Das S: Flowering time control: gene network modelling and the link to quantitative genetics. AustralianJournal of Agricultural Research. 2005, 56: 919-936. 10.1071/AR05155. 10.1071/AR05155View ArticleGoogle Scholar
- Gjuvsland AB, Hayes BJ, Omholt SW, Carlborg O: Statistical epistasis is a generic feature of gene regulatory networks. Genetics. 2007, 175: 411-420. 10.1534/genetics.106.058859PubMed CentralPubMedView ArticleGoogle Scholar
- Hansen TF, Wagner GP: Modeling genetic architecture: a multilinear theory of gene interaction. Theoretical Population Biology. 2001, 59: 61-86. 10.1006/tpbi.2000.1508PubMedView ArticleGoogle Scholar
- Bintu L, Buchler NE, Garcia HG, Gerland U, Hwa T, Kondev J, Kuhlman T, Phillips R: Transcriptional regulation by the numbers: applications. Current Opinion in Genetics & Development. 2005, 15: 125-135. 10.1016/j.gde.2005.02.006View ArticleGoogle Scholar
- Bintu L, Buchler NE, Garcia HG, Gerland U, Hwa T, Kondev J, Phillips R: Transcriptional regulation by the numbers: models. Current opinion in genetics & development. 2005, 15: 116-124. 10.1016/j.gde.2005.02.007View ArticleGoogle Scholar
- Buchler NE, Gerland U, Hwa T: On schemes of combinatorial transcription logic. Proc Natl Acad Sci U S A. 2003, 100 (9): 5136-5141. 10.1073/pnas.0930314100PubMed CentralPubMedView ArticleGoogle Scholar
- Gehring NH, Frede U, Neu-Yilik G, Hundsdoerfer P, Vetter B, Hentze MW, Kulozik AE: Increased efficiency of mRNA 3' end formation: a new genetic mechanism contributing to hereditary thrombophilia. Nature Genetics. 2001, 28: 389-392. 10.1038/ng578PubMedView ArticleGoogle Scholar
- Hoogendoorn B, Coleman SL, Guy CA, Smith K, Bowen T, Buckland PR, O'Donovan MC: Functional analysis of human promoter polymorphisms. Human Molecular Genetics. 2003, 12: 2249-2254. 10.1093/hmg/ddg246PubMedView ArticleGoogle Scholar
- Peng J, Murray EL, Schoenberg DR: The poly(A)-limiting element enhances mRNA accumulation by increasing the efficiency of pre-mRNA 3' processing. RNA. 2005, 11: 958-965. 10.1261/rna.2020805PubMed CentralPubMedView ArticleGoogle Scholar
- Rosenfeld N, Young JW, Alon U, Swain PS, Elowitz MB: Gene regulation at the single-cell level. Science. 2005, 307: 1962-1965. 10.1126/science.1106914PubMedView ArticleGoogle Scholar
- Carey M, Smale ST: Transcriptional regulation in eukaryotes: concepts, strategies, and techniques. 2000, Cold Spring Harbor, NY: Cold Spring Harbor Laboratory PressGoogle Scholar
- Lewin B: Genes VIII. 2004, Upper Saddle River, N.J.: Pearson Prentice HallGoogle Scholar
- Shen-Orr SS, Milo R, Mangan S, Alon U: Network motifs in the transcriptional regulation network of Escherichia coli. Nature Genetics. 2002, 31: 64-68. 10.1038/ng881PubMedView ArticleGoogle Scholar
- Lee TI, Rinaldi NJ, Robert F, Odom DT, Bar-Joseph Z, Gerber GK, Hannett NM, Harbison CT, Thompson CM, Simon I: Transcriptional regulatory networks in Saccharomyces cerevisiae. Science. 2002, 298: 799-804. 10.1126/science.1075090PubMedView ArticleGoogle Scholar
- Becskei A, Serrano L: Engineering stability in gene networks by autoregulation. Nature. 2000, 405: 590-593. 10.1038/35014651PubMedView ArticleGoogle Scholar
- Snoussi E, Thomas R: Logical Identification of all Steady-States: the Concept of Feedback Loop Characteristic States. Bulletin of Mathematical Biology. 1993, 55: 973-991.View ArticleGoogle Scholar
- Mayo AE, Setty Y, Shavit S, Zaslaver A, Alon U: Plasticity of the cis-regulatory input function of a gene. PLoS Biology. 2006, 4: e45- 10.1371/journal.pbio.0040045PubMed CentralPubMedView ArticleGoogle Scholar
- Davidson EH: The Regulatory Genome: Gene Regulatory Networks In Development And Evolution. 2006, San Diego, CA: Academic PressGoogle Scholar
- Wang J, Ellwood K, Lehman A, Carey MF, She ZS: A mathematical model for synergistic eukaryotic gene activation. Journal of Molecular Biology. 1999, 286: 315-325. 10.1006/jmbi.1998.2489PubMedView ArticleGoogle Scholar
- Hooshangi S, Thiberge S, Weiss R: Ultrasensitivity and noise propagation in a synthetic transcriptional cascade. Proc Natl Acad Sci U S A. 2005, 102: 3581-3586. 10.1073/pnas.0408507102PubMed CentralPubMedView ArticleGoogle Scholar
- Kringstein AM, Rossi FM, Hofmann A, Blau HM: Graded transcriptional response to different concentrations of a single transactivator. Proc Natl Acad Sci U S A. 1998, 95: 13670-13675. 10.1073/pnas.95.23.13670PubMed CentralPubMedView ArticleGoogle Scholar
- Veitia RA: A sigmoidal transcriptional response: cooperativity, synergy and dosage effects. Biological Reviews. 2003, 78: 149-170. 10.1017/S1464793102006036PubMedView ArticleGoogle Scholar
- Savageau MA: Michaelis-Menten mechanism reconsidered: implications of fractal kinetics. Journal of Theoretical Biology. 1995, 176: 115-124. 10.1006/jtbi.1995.0181PubMedView ArticleGoogle Scholar
- Verma M, Rawool S, Bhat PJ, Venkatesh KV: Biological significance of autoregulation through steady state analysis of genetic networks. Bio Systems. 2006, 84: 39-48.PubMedView ArticleGoogle Scholar
- Falconer DS, Mackay TFC: Introduction to quantitative genetics. 1996, Harlow: Longman GroupGoogle Scholar
- Schadt EE, Monks SA, Drake TA, Lusis AJ, Che N, Colinayo V, Ruff TG, Milligan SB, Lamb JR, Cavet G: Genetics of gene expression surveyed in maize, mouse and man. Nature. 2003, 422: 297-302. 10.1038/nature01434PubMedView ArticleGoogle Scholar
- Bergman A, Siegal ML: Evolutionary capacitance as a general feature of complex gene networks. Nature. 2003, 424: 549-552. 10.1038/nature01765PubMedView ArticleGoogle Scholar
- Siegal ML, Bergman A: Waddington's canalization revisited: developmental stability and evolution. Proceedings of the National Academy of Sciences of the United States of America. 2002, 99: 10528-10532. 10.1073/pnas.102303999PubMed CentralPubMedView ArticleGoogle Scholar
- Lynch M, Walsh B: Genetics and analysis of quantitative traits. 1998, Sunderland, Mass: Sinauer AssociatesGoogle Scholar
- Hudson TJ: Wanted: regulatory SNPs. Nature Genetics. 2003, 33: 439-440. 10.1038/ng0403-439PubMedView ArticleGoogle Scholar
- Yan H, Yuan W, Velculescu VE, Vogelstein B, Kinzler KW: Allelic variation in human gene expression. Science. 2002, 297: 1143- 10.1126/science.1072545PubMedView ArticleGoogle Scholar
- Doss S, Schadt EE, Drake TA, Lusis AJ: Cis-acting expression quantitative trait loci in mice. Genome Research. 2005, 15: 681-691. 10.1101/gr.3216905PubMed CentralPubMedView ArticleGoogle Scholar
- Ronald J, Brem RB, Whittle J, Kruglyak L: Local Regulatory Variation in Saccharomyces cerevisiae. PLoS Genetics. 2005, 1: e25- 10.1371/journal.pgen.0010025PubMed CentralPubMedView ArticleGoogle Scholar
- Lipshtat A, Perets HB, Balaban NQ, Biham O: Modeling of negative autoregulated genetic networks in single cells. Gene. 2005, 347: 265-271. 10.1016/j.gene.2004.12.016PubMedView ArticleGoogle Scholar
- Haley CS, Knott SA: A Simple Regression Method for Mapping Quantitative Trait Loci in Line Crosses Using Flanking Markers. Heredity. 1992, 69: 315-324.PubMedView ArticleGoogle Scholar
- Jiang C, Zeng ZB: Multiple trait analysis of genetic mapping for quantitative trait loci. Genetics. 1995, 140: 1111-1127.PubMed CentralPubMedGoogle Scholar
- Lander ES, Botstein D: Mapping Mendelian Factors Underlying Quantitative Traits Using Rflp Linkage Maps. Genetics. 1989, 121: 185-199.PubMed CentralPubMedGoogle Scholar
- Grigull J, Mnaimneh S, Pootoolal J, Robinson MD, Hughes TR: Genome-wide analysis of mRNA stability using transcription inhibitors and microarrays reveals posttranscriptional control of ribosome biogenesis factors. Mol Cell Biol. 2004, 24 (12): 5534-5547. 10.1128/MCB.24.12.5534-5547.2004PubMed CentralPubMedView ArticleGoogle Scholar
- Raghavan A, Ogilvie RL, Reilly C, Abelson ML, Raghavan S, Vasdewani J, Krathwohl M, Bohjanen PR: Genome-wide analysis of mRNA decay in resting and activated primary human T lymphocytes. Nucleic Acids Research. 2002, 30: 5529-5538. 10.1093/nar/gkf682PubMed CentralPubMedView ArticleGoogle Scholar
- Hunter PJ, Borg TK: Integration from proteins to organs: the Physiome Project. Nature Reviews Molecular Cell Biology. 2003, 4: 237-243. 10.1038/nrm1054PubMedView ArticleGoogle Scholar
- Mestl T, Plahte E, Omholt SW: A mathematical framework for describing and analysing gene regulatory networks. Journal of Theoretical Biology. 1995, 176: 291-300. 10.1006/jtbi.1995.0199PubMedView ArticleGoogle Scholar
- Plahte E, Mestl T, Omholt SW: A methodological basis for description and analysis of systems with complex switch-like interactions. Journal of Mathematical Biology. 1998, 36: 321-348. 10.1007/s002850050103PubMedView ArticleGoogle Scholar
- Hill AV: The possible effect of the aggregation of the molecules of hemoglobin. Journal of Physiology. 1910, 40: IV-VIII.Google Scholar
- Broman KW, Wu H, Sen S, Churchill GA: R/qtl: QTL mapping in experimental crosses. Bioinformatics. 2003, 19: 889- 10.1093/bioinformatics/btg112PubMedView ArticleGoogle Scholar
- Basten CJ, Weir BS, Zeng Z-B: QTL Cartographer, Version 1.16. 2002, Raleigh, NC.: Department of Statistics, North Carolina State UniversityGoogle Scholar
- Basten CJ, Weir BS, Zeng ZB: Zmap a QTL cartographer. Proceedings of the 5th World Congress on Genetics Applied to Livestock Production: Computing Strategies and Software; 1994; Guelph, Ontario, Canada. 1994, 65-66.Google Scholar
- Cheverud JM, Routman EJ: Epistasis and its contribution to genetic variance components. Genetics. 1995, 139: 1455-1461.PubMed CentralPubMedGoogle 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.