Mixed Model Methodology Research Articles

Pharmacokinetics

AbstractPharmacokinetics is the study of the absorption, distribution, metabolism, and excretion of drugs. Simply put, pharmacokinetics is what the body does to the drug, which is opposed to pharmacodynamics, which is what the drug does to the body. This review will introduce pharmacokinetics as a science showing how biochemistry, biology, pharmacology, and physiology interact to explain how a drug ‘works’. Pharmacokinetics is becoming increasingly quantitative with population pharmacokinetics–pharmacodynamics, which uses nonlinear mixed effects model methodology, becoming more and more commonplace and acting as the core to model‐based drug development. This review will show how statistics is becoming increasingly more important in pharmacokinetic analysis and provide an introduction to statistical analysis of pharmacokinetic data. WIREs Comp Stat 2011 3 332–342 DOI: 10.1002/wics.153This article is categorized under: Applications of Computational Statistics > Computational and Molecular Biology

Analysis of natural allelic variation in Arabidopsis using a multiparent recombinant inbred line population

To exploit the diversity in Arabidopsis thaliana, eight founder accessions were crossed to produce six recombinant inbred line (RIL) subpopulations, together called an Arabidopsis multiparent RIL (AMPRIL) population. Founders were crossed pairwise to produce four F1 hybrids. These F1s were crossed according to a diallel scheme. The resulting offspring was then selfed for three generations. The F4 generation was genotyped with SNP and microsatellite markers. Data for flowering time and leaf morphology traits were determined in the F5 generation. Quantitative trait locus (QTL) analysis for these traits was performed using especially developed mixed-model methodology, allowing tests for QTL main effects, QTL by background interactions, and QTL by QTL interactions. Because RILs were genotyped in the F4 generation and phenotyped in the F5 generation, residual heterozygosity could be used to confirm and fine-map a number of the QTLs in the selfed progeny of lines containing such heterozygosity. The AMPRIL population is an attractive resource for the study of complex traits.

Desempenho genotípico de linhagens de arroz de terras altas utilizando metodologia de modelos mistos

Avaliou-se o desempenho de genótipos de arroz de terras altas utilizando metodologia de modelos mistos. Foram utilizados dados de produtividade de grãos dos ensaios de valor de cultivo e uso (VCU) do programa cooperativo de melhoramento de arroz de terras altas, desenvolvido em Minas Gerais, no período de 1997/1998 a 2007/2008. Realizou-se análise de deviance para verificar a significância dos efeitos de genótipos e suas interações com locais e anos, obtendo-se a estabilidade e a adaptabilidade genotípicas, com os métodos da média harmônica dos valores genotípicos. Os efeitos de genótipos, genótipos x locais e genótipos x anos foram estatisticamente significativos (p<0,05). As linhagens e cultivares mais estáveis foram: Curinga-3, CNA 8983, Guarani, BRSMG Caravera e CNA 8824. Com maior adaptabilidade, estabilidade e produtividade de grãos, simultaneamente, destacaram-se a BRSMG Caravera, Curinga-3, MG 1089, MG 1097 e CNA 8436. A BRSMG Caravera, entre todas as cultivares e linhagens avaliadas de 1997 a 2008, foi a de melhor desempenho pela MHPRVG, ou seja, maior estabilidade, adaptabilidade e produtividade de grãos, simultaneamente, proporcionando aos orizicultores de Minas Gerais alta segurança no seu cultivo.

Effects of Presorting on Stress Responses at Loading and Unloading and the Impact on Transport Losses from Market-Weight Pigs

The objective of this study was to determine the effects of presorting before loading on stress responses and transport losses at the slaughter facility in the market-weight pig. A total of 5,802 pigs (n = 33 loads) were used in a randomized complete block design with 2 treatments. The presorted (PRE) and not presorted (NON) treatments each had 292 pigs/pen (0.65 m2/pig). For the PRE treatment, internal swing gates were used to presort pigs manually 18 h before loading, whereas pigs in the NON treatment were sorted from pen mates at the time of marketing. Treatments were assigned to trailer decks in an alternating manner. Data were analyzed using generalized linear mixed model methodology, and loading time was analyzed using mixed model procedures. Loading time differed (P 0.05) different. There were no (P > 0.05) differences between treatments for total losses at the slaughter facility. Two pigs in the PRE treatment and zero pigs in the NON treatment were classified as dead on arrival. In conclusion, presorting market-weight pigs reduced loading time and some stress responses on farm; however, no treatment differences were observed for stress responses or transport losses at the slaughter facility.

Methods of plant breeding in the genome era

Methods of genomic value prediction are reviewed. The majority of the methods are related to mixed model methodology, either explicitly or implicitly, by treating systematic environmental effects as fixed and quantitative trait locus (QTL) effects as random. Six different methods are reviewed, including least squares (LS), ridge regression, Bayesian shrinkage, least absolute shrinkage and selection operator (Lasso), empirical Bayes and partial least squares (PLS). The LS and PLS methods are non-Bayesian because they do not require probability distributions for the data. The PLS method is introduced as a special dimension reduction scheme to handle high-density marker information. Theory and methods of cross-validation are described. The leave-one-out cross-validation approach is recommended for model validation. A working example is used to demonstrate the utility of genome selection (GS) in barley. The data set contained 150 double haploid lines and 495 DNA markers covering the entire barley genome, with an average marker interval of 2·23 cM. Eight quantitative traits were included in the analysis. GS using the empirical Bayesian method showed high predictability of the markers for all eight traits with a mean accuracy of prediction of 0·70. With traditional marker-assisted selection (MAS), the average accuracy of prediction was 0·59, giving an average gain of GS over MAS of 0·11. This study provided strong evidence that GS using marker information alone can be an efficient tool for plant breeding.

Towards understanding and use of mixed-model analysis of agricultural experiments

Despite the presence of both fixed and random effects in most agricultural experiments, many crop researchers have continued use of the conventional analysis of variance (ANOVA) model or general linear model (GLM) that provides a correct analysis only if all the effects are fixed. Ignoring or mistreating random effects may have inadvertently led to inappropriate analyses and thus to dubious conclusions appearing in the scientific literature. The objective of this paper is to provide a tutorial account of the mixed-model methodology and its applications to the analysis of agricultural experiments. The description and discussion on ANOVA vs. mixed-effect models center on the use of PROC GLM vs. PROC MIXED of the SAS® System. This paper points out the need for mixed-model analysis, describes and discusses key new features and properties of mixed-model analysis that would facilitate the understanding and use of PROC MIXED. Additionally, it analyzes and interprets three examples: comparison between two samples, and analyses of randomized complete design and split-plot design. Appendices include SAS code and theory underlying mixed-model analysis which will help gain hands-on experiences and ensure correct interpretation of SAS outputs by PROC MIXED. Such a comparative assessment of GLM vs. MIXED procedures will help to underscore the key advantages of PROC MIXED and to convince GLM users to make a true transition towards the increased and appropriate use of PROC MIXED in agricultural experiments.Key words: Analysis of variance, fixed vs. random effects, general linear models, inference spaces, mixed models, randomized complete block design, split-plot design

Updated estimates of HAL n and RN− effects on pork quality: Fresh and processed loin and ham

A 1000-pig F2 intercross QTL detection experimental population was generated using two commercial sire lines. Independent carriers of HAL n and RN- mutations (10% and 14%, respectively) were included in this population as control genotypes. The effects of HAL n and RN- heterozygous genotypes on fresh and transformed loins and hams were estimated using a mixed model methodology. The results document the unfavorable effects of both mutations on meat quality. Smaller effects of HAL Nn genotype compared to HAL nn or RN-rn+ genotypes were estimated. Interestingly, effects of HAL Nn genotype on meat pH and loin color could be insignificant at 24-h postmortem, but translate into higher water losses on storage and cooking, and result in tougher cooked loin. Using the same methodology, significant effects of the PRKAG3 (RN) I199 allele on ultimate pH values but not on glycolytic potential were observed.

GENETIC RESPONSE TO SELECTION FOR REPRODUCTIVE PERFORMANCE IN A MATERNAL LINE OF RABBITS""

An evaluation of genetic response on reproductive traits was carried out in a line of rabbits selected for litter size at weaning. The evaluation was performed comparing contemporary does of generation 17 and 26, using cryopreservation techniques. The direct response to selection between both generations was 0.77±0.27 rabbits weaned/litter and there were favorable correlated responses in total litter size at birth and born alive. There was a non-significant response in prenatal traits, but the litter size results imply response probably in foetal survival. The genetic trends were also estimated using mixed model methodology and were higher than the previous ones. Therefore, the repeatability animal model and the genetic parameters used are unsuitable. Models taking into account effects such as dominance need to be investigated.

Genetic trend in selection for litter weight in two maternal lines of rabbits in Egypt.

An analysis was carried out to evaluate the results of the selection program of a new synthetic maternal line of Egyptian APRI rabbits and the Spanish V line. The selection criterion was litter weaning weight in both lines based on Best Linear Unbiased Prediction (BLUP) estimations under a repeatability animal model. The studied traits were: litter birth weight, litter weaning weight (at 28 d) and pre-weaning litter gain. Heritability estimates were generally low (0.09 to 0.11) and comparable in both lines. Low to moderate estimates of repeatability were observed for all studied traits (0.15 to 0.31). Estimates of genetic correlations were high (0.77 to 0.97), while permanent environmental effects correlations were mostly lower than genetic correlations (0.47 to 0.87). The genetic trends were also estimated using mixed model methodology and were significant and comparable (34.2 and 32.5 g) for the selected trait (litter weaning weight) in APRI and V lines, respectively. Furthermore, correlated genetic trends were significant (P

Continuous Time Brownian Motion Models for Analysis of Sequential Data

SummaryThe paper discusses techniques for analysis of sequential data from variable processes, particularly techniques that can be used even when the data are not equally spaced in time. The techniques are based on models which use continuous time Brownian motion and its integrals to describe the pattern of variation in an underlying physical process and use white noise to describe variation due to measurement processes. The paper concentrates on making statements about what has been happening in the region that is covered by the data rather than on making predictions about regions that are far from the data. It is argued that the continuous time integrated Brownian motion plus white noise model is always preferable to its discrete time analogue: the local linear trend model. Further integrals of Brownian motion can be fitted as an alternative to using splines. All of these continuous time Brownian motion models can be fitted to data that are associated with a time interval (interval data) as well as to data that are associated with a single time (spot data). They can be fitted by using mixed model methodology as well as by using Kalman filtering and smoothing.

Effects of Facility System Design on the Stress Responses and Market Losses of Market Weight Pigs During Loading and Unloading

<h2>ABSTRACT</h2> The objective of this study was to determine the effect of 2 grow-finish swine facility system designs and presorting on the stress responses and incidence of nonambulatory and dead pigs at the time of loading and unloading market pigs. A total of 5,901 pigs (n=33 loads) were used in a randomized complete block design. The new facility system design (NFSD) had 192 pigs/pen, and internal swing gates were used to manually presort market weight pigs on the day before loading. The smaller traditional facility system design (TFSD) had 32 pigs/pen with no presorting. During loading, facility system design and presorting were randomly assigned to trailer decks. The time required to load a trailer deck was recorded by facility system design and presorting. Data were analyzed using generalized linear mixed model methodology, and loading time was analyzed using mixed model procedures. Loading time differed (P=0.007) between facility system design and presorting, with TFSD taking longer to load compared with NFSD. The NFSD with presorting had lower (P<0.05) percentages of open-mouth breathing, skin discoloration, and muscle tremors during loading and unloading compared with TFSD with no presorting. The NFSD had fewer (P=0.006) dead pigs on arrival and fewer-total nonambulatory pigs (P=0.041) and total losses (P=0.003) at the plant compared with TFSD. In summary, the NFSD with presorting reduced physical signs of stress during loading and unloading and reduced total losses at the plant by 66% compared with losses attained when pigs were loaded from the TFSD with no presorting.

Goat breeding research in Mexico

Goat breeding research in Mexico began in 1980 with studies on breed and crossbreed comparisons involving local goats and goats from specialized breeds imported from the USA, using data from a large goat breeding centre, in Tlahualilo, Durango, located in Northern Mexico. Genetic groups were purebred and crossbreed goats involving recurrent backcrosses with imported breeds and local Granadina breed with nondescript local goats. Traits analyzed include milk production, kidding intervals, kidding distributions, kid weights, survival and stayability. Results showed the higher milk production efficiency of Alpine, Saanen and Toggenburg, compared to local populations, but there were smaller advantages from the use of Nubian or Granadina goats. Granadina and Nubian goats had higher kid survival rates, longer breeding seasons and were more efficient for meat production. Evidence for higher stayability for adult Alpine and lower for Saanen goats was reported. More recently, research on meat goat breeding involving local goats and other genetic groups such as Boer, Nubian and breed crosses have been done. Research on genetic parameter estimation began with estimates of heritability for birth weight in kids and repeatabilities for several milk production and reproduction traits, and has continued until now. It is currently related to milk production and composition traits recorded in dairy breeding populations of Alpine, Saanen and Toggenburg goats in the State of Guanajuato. These parameters are being used together with mixed-model methodology for the genetic evaluation of these populations for milk, fat and protein production. Recent studies (2006) regarding the distribution of polymorphism for the α-s1 locus in several goat populations in Mexico and its effects on milk traits in Mexican dairy goats were conducted.

The effect of missing marker genotypes on the accuracy of gene-assisted breeding value estimation: a comparison of methods

In livestock populations, missing genotypes on a large proportion of the animals is a major problem when implementing gene-assisted breeding value estimation for genes with known effect. The objective of this study was to compare different methods to deal with missing genotypes on accuracy of gene-assisted breeding value estimation for identified bi-allelic genes using Monte Carlo simulation. A nested full-sib half-sib structure was simulated with a mixed inheritance model with one bi-allelic quantitative trait loci (QTL) and a polygenic effect due to infinite number of polygenes. The effect of the QTL was included in gene-assisted BLUP either by random regression on predicted gene content, i.e. the number of positive alleles, or including haplotype effects in the model with an inverse IBD matrix to account for identity-by-descent relationships between haplotypes using linkage analysis information (IBD–LA). The inverse IBD matrix was constructed using segregation indicator probabilities obtained from multiple marker iterative peeling. Gene contents for unknown genotypes were predicted using either multiple marker iterative peeling or mixed model methodology. For both methods, gene-assisted breeding value estimation increased accuracies of total estimated breeding value (EBV) with 0% to 22% for genotyped animals in comparison to conventional breeding value estimation. For animals that were not genotyped, the increase in accuracy was much lower (0% to 5%), but still substantial when the heritability was 0.1 and when the QTL explained at least 15% of the genetic variance. Regression on predicted gene content yielded higher accuracies than IBD–LA. Allele substitution effects were, however, overestimated, especially when only sires and males in the last generation were genotyped. For juveniles without phenotypic records and traits measured only on females, the superiority of regression on gene content over IBD–LA was larger than when all animals had phenotypes. Missing gene contents were predicted with higher accuracy using multiple-marker iterative peeling than with using mixed model methodology, but the difference in accuracy of total EBV was negligible and mixed model methodology was computationally much faster than multiple iterative peeling. For large livestock populations it can be concluded that gene-assisted breeding value estimation can be practically best performed by regression on gene contents, using mixed model methodology to predict missing marker genotypes, combining phenotypic information of genotyped and ungenotyped animals in one evaluation. This technique would be, in principle, also feasible for genomic selection. It is expected that genomic selection for ungenotyped animals using predicted single nucleotide polymorphism gene contents might be beneficial especially for low heritable traits.

Correlative responses for carcass and meat quality traits to selection for ovulation rate or prenatal survival in French Large White pigs1

Correlated effects of selection for components of litter size on carcass and meat quality traits were estimated using data from 3 lines of pigs derived from the same Large White base population. Two lines were selected for 6 generations on high ovulation rate at puberty (OR) or high prenatal survival corrected for ovulation rate in the first 2 parities (PS). The third line was an unselected control (CON). The 3 lines were kept for a 7th generation, but without any selection. Carcass and meat quality traits were recorded on the 5th to 7th generation of the experiment. Carcass traits included dressing percentage, carcass length (LGTH), average backfat thickness (ABT), estimated lean meat content, and 8 carcass joint weight traits. Meat quality traits included pH recorded 24 h after slaughter (pH24) of LM, gluteus superficialis (GS), biceps femoris (BF), and adductor femoris (AD) muscles, as well as reflectance and water-holding capacity (WHC) of GS and BF muscles. Heritabilities of carcass and meat quality traits and their genetic correlations with OR and PS were estimated using REML methodology applied to a multiple trait animal model. Correlated responses to selection were then estimated by computing differences between OR or PS and CON lines at generations 5 to 7 using least squares and mixed model methodology. Heritability (h(2)) estimates were 0.08 +/- 0.04, 0.58 +/- 0.10, 0.70 +/- 0.10, and 0.74 +/- 0.10 for dressing percentage, LGTH, ABT, and lean meat content, respectively, ranged from 0.28 to 0.72 for carcass joint traits, from 0.28 to 0.45 for pH24 and reflectance measurements, and from 0.03 to 0.11 for WHC measurements. Both OR and PS had weak genetic correlations with carcass (r(G) = -0.09 to 0.17) and most meat quality traits. Selection for OR did not affect any carcass composition or meat quality trait. Correlated responses to selection for PS were also limited, with the exception of a decrease in pH24 of GS and BF muscles (-0.12 to -0.14 after 6 generations; P < 0.05), in WHC of GS muscle (-18.9 s after 6 generations; P < 0.05) and a tendency toward an increase in loin weight (0.44 kg after 6 generations; P < 0.10) .

Variability of Urinary Metabolite Levels of Pesticides in Turf Applicators; Implications for Exposure Measurement in Prospective Epidemiologic Studies

ISEE-0030 Background and Objectives: Epidemiologic studies of pesticides have been subject to important biases arising from exposure misclassification. While turf applicators are exposed to a variety of herbicides and insecticides, these exposures have not been well characterized. Methods: To examine variations in urinary markers of pesticides, repeated urine samples were obtained from 135 TruGreen turf applicators employed in six branches across the U.S. Urine samples (n = 1028) were collected in the 3 spraying seasons: the spring (April and May) and fall (October and November) herbicide sprays, as well as the summertime (June and July) insecticide spray. Total urine was collected for two consecutive 24-hour periods during the herbicide sprays, and four consecutive 12-hour periods (insecticide) following a minimum of 3 consecutive workdays. Measured pesticides included: 2,4-D, MCPA, Mecoprop, dicamba, the bifenthrin metabolite (MPA) and imidacloprid, and its metabolite 6-CNA. High performance liquid chromatography tandem mass spectrometry was used to measure ppb levels of pesticides in the 12 and 24 hour samples and these were used to estimate total amount of pesticides in μg over a 24 hour period. The distributions of these exposures were characterized through the use of boxplots. Mixed model methodology was employed to between and within worker variability. Results: The intraclass correlation coefficients (ICC) for all pesticides were less than 0.31, indicating considerable within subject variability over the three spraying seasons. However, considerably higher ICCs were observed when estimated separately within each season (ICC for MCPA overall was 0.31 but ranged from 0.63 to 0.89 when analysed by season). Conclusions: Exposure measurement error could be reduced by performing sampling within individuals across seasons for the majority of herbicides tested. For some pesticides (i.e. MCPP and imidacloprid), repeated samples within seasons may be necessary to characterize exposures. The implications of these results within the context of a prospective cohort study are discussed.

Short-term selection for yearling weight in a small-experimental Iranian Afshari sheep flock

A selection experiment was initiated in 2000 in an Afshari sheep flock at the department of animal breeding and genetics of the University of Zanjan, Iran. The aim was to evaluate the response of Afshari sheep to selection for yearling live weight. Here, we evaluate the results of this breeding program obtained between 2000 and 2005. Traits studied were birth weight (BW), weaning weight (WW), yearling weight (YW), average daily gain from birth to weaning (WWDG) and average daily gain from weaning to yearling age (YWDG). Mixed model methodology based on a multi-trait animal model was employed to estimate (co)variance components and corresponding genetic parameters. Estimates of breeding values were obtained by the best linear unbiased prediction (BLUP) method. Generation intervals on the four genetic pathways were estimated as the average age of parents at the birth of their progeny kept for reproduction. The heritability estimates were 0.34, 0.27, 0.14, 0.20 and 0.01 for BW, WW, YW, WWDG and YWDG, respectively. Estimates of genetic correlations among traits studied were positive, and ranged from low (0.07, YW/WWDG) to high (0.76, YW/YWDG). Genetic improvements over the experiment based on estimated breeding values were 0.104, 0.824, 1.247, 0.005 and ≈0.00 kg for BW, WW, YW, WWDG and YWDG, respectively. Annual genetic gain for YW was relatively high, 0.311 kg yr-1, which demonstrated the effectiveness of the implemented breeding program. Correlated responses in BW, WW, WWDG and YWDG were 0.021, 0.167, 0.001 and ≈0.00 kg yr-1, respectively. Estimates of heritabilities and observed genetic trends indicated that selective breeding can lead to significant genetic improvement in Afshari sheep. The average generation interval was estimated to be 3.35 yr. The shorter generation interval was observed on the sire side compared with the dam side (3.30 yr vs. 3.78 yr), indicating faster generation turnover for sires than for dams. Key words: Sheep, animal model, genetic trend, generation interval, heritability

A model to predict 24-h urinary creatinine using repeated measurements in an occupational cohort study

Creatinine measurements can be used to standardize urinary pesticide concentrations and to estimate "completeness" of urine collections. Published statistical models exist to predict 24-h creatinine, but many were developed assuming independence among observations. Using correlated repeated measurement data collected from an occupational cohort, the objectives were to create a predictive model for 24-h urinary creatinine and to compare the predictive capability of this model to earlier published models. Using a mixed-model methodology, the appropriate covariance structure was identified and utilized to model the measurements. A backwards elimination model building technique applied to the model building data set (110 adult male subjects and 457 creatinine values) yielded a final model that included variables for body mass index (BMI), height, diabetes, allergies, medical conditions that affect kidney function, use of creatine supplements, and anti-inflammatory medications. Using an external model validation data set (21 adult male subjects' creatinine values, n=91 observations from a total of 275) the predictive performance of the model was evaluated using the mean square prediction error (MSPR) and the Pearson's correlation coefficient (r); its performance was better (MSPR=279184, r=0.43) than any of the earlier models investigated (MSPR: range 658860-393139; r, range 0.18-0.38). In conclusion, the use of a covariance structure that allowed repeated measurements for any one individual to be correlated, improved the predictive performance. For purposes of incomplete urine sample identification in observational studies, it is necessary to collect information in addition to age, gender, and BMI, which are typically used in these settings.

Selection of alfalfa cultivars adapted for tropical environments with repeated measures using PROC MIXED of SAS® System

Although alfalfa (Medicago sativa L.) is a leguminous herbage widely used in temperate regions as animal feed, there is not much research in tropical regions to develop cultivars adapted to these environmental conditions. The utilization of adapted cultivars with adequate management practices is important to improve productivity, quality and persistence of cultivated pastures. The objectives of this study were to verify the genetic variability among alfalfa cultivars and to rank them using mixed model methodology. A total of 35 alfalfa cultivars were evaluated in the rainy and dry seasons, from 1996 to 2000, in plots of 2.8 m2 in Sertãozinho, São Paulo, Brazil. The experimental design was a randomized complete block with three replications. Longitudinal data of dry matter yield were analyzed using PROC MIXED of SAS® System. Several covariance structures were tested and the spherical spatial structure was selected. The results show that the genetic variability was statistically significant only for the dry season. Moreover, the interaction among cultivars and harvests variance was highly significant for both seasons. The empirical best linear unbiased predictions of cultivar effects were obtained, allowing for the selection of the superior cultivars MH 15, 5715, SW 8210, Rio, High, 5888, Monarca, Victoria, Florida 77 and Falcon. Crioula, the most common cultivar in Brazil, showed low forage potential in Sertãozinho. Results indicate potential for use of more productive cultivars of alfalfa to produce animal feed in tropical environments.

High School Social Studies Teachers’ Attitudes Toward English Language Learners

This study examined the attitudes of high school social studies teachers toward the inclusion of English language learners (ELLs) in their social studies classrooms. A mixed-model methodology was employed by surveying all 344 social studies teachers in the school district. Additionally, eight of these teachers were interviewed to ascertain their attitudes toward including ELLs in high school social studies classrooms. More than three-fourths of teachers surveyed indicated they would prefer that ELLs not be in their classroom until they have “learned” English. Furthermore, the main cause of the negative attitudes seemed to be rooted in the inability of the social studies teachers to effectively modify instruction. The possible results of these attitudes and the efficacy of current policies are discussed.

A powerful and flexible linear mixed model framework for the analysis of relative quantification RT-PCR data

Quantitative reverse transcription polymerase chain reaction (qRT-PCR) is currently viewed as the most precise technique to quantify levels of messenger RNA. Relative quantification compares the expression of a target gene under two or more experimental conditions normalized to the measured expression of a control gene. The statistical methods and software currently available for the analysis of relative quantification of RT-PCR data lack the flexibility and statistical properties to produce valid inferences in a wide range of experimental situations. In this paper we present a novel method for the analysis of relative quantification of qRT-PCR data, which consists of the analysis of cycles to threshold values ( C T) for a target and a control gene using a general linear mixed model methodology. Our method allows testing of a broader class of hypotheses than traditional analyses such as the classical comparative C T. Moreover, a simulation study using plasmode datasets indicated that the estimated fold-change in pairwise comparisons was the same using either linear mixed models or a comparative C T method, but the linear mixed model approach was more powerful. In summary, the method presented in this paper is more accurate, powerful and flexible than the traditional methods for analysis of qRT-PCR data. This new method is especially useful for studies involving multiple experimental factors and complex designs.