Genetic component of sensorimotor capacity in rats
Lauren Gerard Koch and
Steven L. Britton
Functional Genomics Laboratory, Medical College of Ohio, Toledo, Ohio 43614-5804
 |
ABSTRACT
|
---|
Creation of generalized genetic models for low and high sensorimotor capacity would be important tools for resolution of this complex trait. As proof-of-principle we estimated phenotypic variation and narrow-sense heritability (h2) of sensorimotor capacity in 19 families of genetically heterogeneous N:NIH rats and in 11 strains of inbred rats. Sensorimotor capacity was defined as the time a rat remained on an accelerating rotorod. N:NIH rats recorded variation in rotorod scores that ranged from 3- to 7-fold. The value of h2, estimated from offspring-parent regression across one generation, was 0.68 for females and 0.74 for males in N:NIH rats. In inbred rats, h2 was estimated by partitioning phenotypic variation into additive genetic and environmental components and averaged 0.39 in females and 0.48 in males. These results demonstrate a heritable component to sensorimotor capacity sufficient for success in developing contrasting genetic models by divergent artificial selection in rats.
inbred strains; N:NIH rats; rotorod; heritability; variation
 |
INTRODUCTION
|
---|
SENSORIMOTOR CAPACITY is a fundamental high-order phenotype that requires integration of multiple systems to execute complicated movements. Although heritability studies demonstrate a significant genetic component to variation in sensorimotor capacity, the allelic variants that contribute to this intrinsic variation remain undefined. In human twins, estimates of the genotypic component of variance in sensorimotor capacity range from 48 to 77% (16, 17, 22, 25, 27, 28). In mice, Swallow (24) reported that narrow-sense heritability (h2) for voluntary wheel running was 0.28. Reed found a broad sense heritability of 0.33 for the delay of an action potential across the neuromuscular junction (21), and an h2 value of 0.23 for peripheral nerve conduction velocity (20). These observations suggest that variance in several likely determinant phenotypes of sensorimotor capacity can be accounted for partially by the summed variance of numerous genetic components.
One reasonable starting point for identifying alleles causative of variation in a complex trait, such as sensorimotor capacity, is to develop genetic models that contrast widely for the trait by artificial divergent selective breeding. Two conditions are ideal for success in selective breeding for a trait. First, wide phenotypic variation should exist in the population for the trait, and second, the trait should demonstrate significant heritability. Therefore, the aim of this study was to test the general hypothesis that phenotypic variation in rotorod capacity has a measurable genetic component as estimated from h2. This information is obligatory and antecedent for the creation of low and high genetic models of this trait by artificial selection.
In previous work, we used the time a rat remained on an accelerating rotorod to estimate the phenotypic variation in sensorimotor capacity among 11 inbred rat strains (2). The rotorod test separated the 11 strains as a wide continuum that ranged from
3-fold differences for female rats to 11-fold differences for male rats. Our experience with the rotorod led us to agree with others (12), that the rotorod can be used as a general test of overall sensorimotor competence.
Thus, for the present study, we applied the accelerating rotorod test to a parental population of N:NIH stock of rats and their offspring to estimate h2 from the regression of offspring on parents. We also estimated h2 by comparing the sensorimotor performance between and within 11 inbred strains. Our results demonstrate that sensorimotor capacity exhibits both wide phenotypic variation and measurable heritability in outcrossed and inbred rats.
 |
Methods
|
---|
All procedures were carried out with the approval of our Institutional Animal Care and Use Committee and were conducted in accordance with the "Guiding Principles in the Care and Use of Animals" as approved by the Council of the American Physiological Society.
Studies in an outbred population.
The parental population was 19 male and 19 female rats from an outbred population of N:NIH stock obtained from a colony maintained at the National Institutes of Health. The N:NIH rats were developed as a genetic resource of a heterogeneous population from the outcross breeding of these eight inbred strains: ACI, BN, BUF, F344, M520, MR, WKY, and WN (9). Each rat in the parental population was of different parentage which broadens the genetic variance (10). The rats were housed two per cage, provided with food (diet 5001; Purina Mills, Richmond, IN) and water ad libitum, and placed on a 12:12-h light-dark cycle with the light cycle coinciding with daytime. The protocol for estimation of sensorimotor capacity with the rotorod was started when the parental rats were 26 wk old.
All rats were tested daily for 5 consecutive days between the hours of 9 AM and noon. The rotorod was built based on the rotating rod apparatus originally described by Dunham and Miya (5). The device consists of a smooth hard plastic cylinder (7.9 cm in diameter and 15 cm long) with concentric 53-cm diameter circular plastic sides attached to prevent the rat from climbing off the cylinder laterally. The cylinder is connected to a variable speed reversible motor, allowing the speed and direction of rotation of the cylinder to be changed. The rat was placed on the cylinder with the long axis of its body parallel to the long axis of the cylinder while the cylinder was rotating at 0.60 rpm. The direction of rotation of the cylinder was alternated every 5 s, and the velocity of rotation was increased every 10 s by 1.3 rpm to a maximum of 48 rpm (two-way rotation test). A padded box was placed 50 cm below the apparatus, so that the rat was not injured when it fell. The score for the test was the number of seconds the rat remained on the rotorod. Rats were not given any positive or negative stimuli from the operator to affect the length of time spent on the rod. Body weight was measured daily after each test session.
The rats were randomly mated, and their offspring were tested on the rotorod at 12 wk of age. By this age, the neurological processes involved in the control of movement are well-developed (1). For the offspring, the test consisted of increasing the velocity of rotation by 1.3 rpm every 5 s, but the direction of rotation was not alternated (one-way rotation test). The parental generation was tested using the two-way rotation procedure because this test showed wider phenotypic variation than the one-way rotation test in our initial study of sensorimotor capacity in inbred rat strains (2). We decided to test the offspring and any subsequent selected generations with the one-way rotation test for four reasons. First, most commercially available rotorods do not have the capability to reverse direction and will limit reproduction of the two-way rotation test for sensorimotor capacity in other laboratories. Second, one-way rotation test had less operator intervention, which will tend to reduce the random error in the measurement of the phenotype. Third, function studies in commercially available inbred mice use one-way rotation that will lend itself for comparison between model systems. Fourth, performance on one-way and two-way rotation tests are closely correlated (R = 0.954) and presumed to test similar function. Figure 1 is a nomogram relating the velocity of rotation (rpm) and distance traversed (cm) to the seconds sustained on the rotorod for both the one-way and two-way rotation tests. The inset to Fig. 1 shows the correlation between scores on the one-way and two-way rotorod tests in 11 inbred strains.

View larger version (27K):
[in this window]
[in a new window]
|
Fig. 1. Nomogram for the one-way and two-way rotation tests on the rotorod showing how distance and speed correspond to the number of seconds attained. Inset: correlation between these two tests among 11 inbred rat strains.
|
|
Narrow-sense heritability (h2) is the extent to which phenotypes in offspring are determined by alleles transmitted from the parents. The value of h2 was estimated for each sex from the slope of the regression of the mean value of offspring (one-way rotation test) vs. the same-sex parent value (two-way rotation test) for the 19 families of N:NIH rats (6). If the trait is additively inherited with complete fidelity, such that the values of offspring are identical to parent, then the slope of the regression line (h2) equals 1. In contrast, if no additive similarity exists between parent and offspring, then h2 equals 0. Finally, each daily rotorod score was regressed linearly on trial numbers (days 15) to test for a possible "learning" effect.
Studies in inbred strains.
Narrow-sense heritability was also estimated from measures of the performance of rats from 11 inbred strains on three tests of sensorimotor capacity, using data from one of our previous studies (2). An inbred strain is defined as a genetically uniform population that was developed by exclusive brother-sister matings for at least 20 generations such that
97.5% of the loci are homozygous. Seven strains (ACI/SegHsd, AUG/OlaHsd, BUF/NHsd, COP/Hsd, DA/OlaHsd, F344/NHsd, and PVG/OlaHsd) were purchased from Harlan Sprague-Dawley (Indianapolis, IN) at 78 wk of age. Four of the strains (LEW, WKY, SR/Jr, and MNS) were obtained from colonies maintained at the Medical College of Ohio (Toledo, OH). For each strain, six rats of each sex were used except as follows: SR/Jr, females, n = 4; AUG/OlaHsd, males, n = 5. None of the strains was originally created by selection for any aspect of sensorimotor capacity.
Sensorimotor capacity was measured with three tests: 1) the one-way rotation test, 2) the two-way rotation test (both of these applied exactly as described above), and 3) a tilt test. For the tilt test, each rat was evaluated for its ability to remain on an inclined platform, a modified version of the test described by Murphy et al. (19). The platform was a rectangular stainless steel pan (24 cm wide x 30 cm long x 15 cm deep) that was set at an initial angle of 22° to the horizontal. At the start of the test the rat was placed in the pan on the side set at 22°, with the long axis of its body parallel to the open edge of the pan. The angle of the pan was then increased by 2° every 5 s to a maximum of 48° at 65 s. The time (s) from when the rat was placed in the pan, until it fell into a padded box, was recorded as the tilt test score.
Narrow-sense heritability (h2) for each of the above three tests was estimated from the ratio of additive genetic variance (VA) to phenotypic variance (VP) (7), with the phenotypic variance being the sum of the VA and environmental variance (VE)
 | (1) |
VA and VE were estimated by partitioning Eq. 1 into the between-group variance component (
B2) and the within-group variance component (
W2) from analysis of variance using these expressions
 | (2) |
 | (3) |
where MSB is the mean square between strains, MSW the mean square within strains, and n is the number of animals in each strain.
B2 equals 2VA, and VE is approximated by
W2. Combining Eqs. 1, 2, and 3, yields an estimate of h2 [Trudy Mackay, personal communication; see also Hegmann and Possidente (11)]
 | (4) |
Because the number of rats was not equal for all strains, a weighted average value for n was calculated as suggested by Sokal and Rohlf (23)
 | (5) |
where a = number of strains, and ni = number of rats in each strain.
Analyses of data.
For each rat, the single best performance out of the five daily rotorod or tilt test trials was used as the measure deemed most closely associated with the genetic component of sensorimotor capacity (2, 14).
Data were analyzed using Sigmastat for Windows Version 2.03 (SPSS Science, Chicago, IL). For the parent and offspring populations, the range of rotorod performance was calculated and the Kolmogorov-Smirnov test was applied to determine whether the data were normally distributed. An F test was applied to test whether the variance was equal in the two sexes. Within each population, the effects of sex and body weight on rotorod performance were determined by one-way analysis of variance. If the data failed the assumption of normality test, then a Kruskal-Wallis one-way analysis of variance on ranks for nonparametric data was applied with a post hoc Dunns multiple comparison procedure. Potential associations between body weight and rotorod scores were assessed by calculating correlation coefficients.
 |
Results
|
---|
Variation in the N:NIH founder population.
Figure 2 shows the best rotorod score for each rat in the parental N:NIH population graphed as seconds on the two-way rotorod test for each sex. The females averaged 70 s on the rotorod, which was not significantly different from the 62 s recorded by the male rats. The data show gradated variation from low to high in rotorod scores that ranged 3.6-fold (35 to 127 s) for the females and 2.7-fold (32 to 86 s) for the males. The scores from the males and females of the parental population were not significantly different from a normal distribution (P > 0.2). On average, the parental population remained on the accelerating rotorod for 65.7 s with a standard deviation of 19.3 s. Comparisons between females and males for body weight and rotorod performance are summarized in Table 1. Body weight was not significantly correlated with rotorod performance in either sex (Pearson correlation coefficients for normally distributed data). There was, however, a significant difference in the variance of sensorimotor capacity (P = 0.04) between the females (variance = 516) and males (variance = 221).

View larger version (15K):
[in this window]
[in a new window]
|
Fig. 2. Distribution of scores for two-way rotation test for the female (n = 19) and male (n = 19) parental rats from the outcrossed N:NIH population. Each point is the best rotorod score for each rat.
|
|
View this table:
[in this window]
[in a new window]
|
Table 1. Comparison between females and males for body weight and rotorod performance in the parental and offspring populations of N:NIH outbred stock
|
|
Variation in the N:NIH offspring.
Figure 3 shows the best rotorod score for each rat in the N:NIH offspring population graphed as seconds on the one-way rotorod test for each sex. The females averaged 62 s on the rotorod, which was significantly greater than the 42 s recorded by the male rats (P > 0.001, Kruskal-Wallis one-way analysis of variance on ranks). The data show gradated variation from low to high in rotorod scores that ranged 7.2-fold (17 to 122 s) for the females and 6.2-fold (15 to 93 s) for the males. The scores from the males (P = 0.003) and females (P = 0.005) of the offspring population were significantly different from a normal distribution (Kolmogorov-Smirnov test). On average, the entire offspring population remained on the accelerating rotorod for 50.3 s with a standard deviation of 22.5 s. For males
4.4% of the variation in performance was negatively associated with body weight (Spearman correlation coefficient for nonnormal data, R = -0.21, P = 0.04). For the females, there was no association between body weight and rotorod score (Table 1).

View larger version (17K):
[in this window]
[in a new window]
|
Fig. 3. Distribution of scores for one-way rotation test for the female (n = 71) and male (n = 92) offspring from the 19 parental pairs shown in Fig. 2. Each point is the best rotorod score for each rat.
|
|
Estimate of heritability in N:NIH rats.
Because the variances were not equal between the sexes, heritability was estimated from the regression of sons on fathers and daughters on mothers (7). Figure 4 shows the regression of the sensorimotor capacity calculated separately for each sex as the mean of the offspring for one-way rotation tests on parental scores for two-way rotation tests. Sensorimotor capacity is represented for 19 families with an average litter size of 8.6 ± 0.7. Figure 4 demonstrates that the parental value was a significant predictor of the sensorimotor capacity of the offspring in both females (R = 0.52, P = 0.023) and males (R = 0.45, P = 0.053). The regression coefficients were 0.34 ± 0.14 for the females and 0.37 ± 0.18 for males. We assume that the regressions of same-sex parents on offspring estimates
VA/VP (i.e., one-half of the additive genetic variance of the parents). Therefore, the regressions and their standard errors were multiplied by two to estimate heritability. The heritability of sensorimotor capacity for the females was 0.68 ± 0.27, and the heritability in the males was 0.74 ± 0.36. These data support the conclusion that
70% of the variation in rotorod performance was accounted for by additive genetic variance.

View larger version (15K):
[in this window]
[in a new window]
|
Fig. 4. Regression of mean offspring score on one parent score for sensorimotor capacity in outbred rats calculated separately for each sex (daughters on mothers and sons on fathers). The slope of the regression line was multiplied by two as an estimate of narrow-sense heritability (h2); h2 in the females was 0.68 ± 0.27, and h2 in the males was 0.74 ± 0.36.
|
|
Possible learning effects in N:NIH rats.
To evaluate whether scores improved with repeated trials, the five daily scores were regressed on the trial number (day of testing) for each sex in both the parental and offspring populations. For both sexes, the parental and offspring rotorod scores correlated positively with trial number (day) (P < 0.001 for all cases, linear model). On average, rats improved their score by 5.5 s per day of testing (Table 2).
Heritability in inbred strains.
For comparative data between inbred strains and sex differences, see previously published article (2). Narrow-sense heritability was estimated for females and males using analysis of variance to derive the between-group variance component (
B2) and the within-group variance component (
W2) as shown above in Eq. 4. Data from which h2 values were estimated are summarized in Table 3.
B2 for the six groups (3 tests, 2 sexes) ranged from 1,084 to 9,952 and all showed significant probabilities that between strain variation existed (P < 5 x 10-8 for all cases). The h2 averaged 0.50 (range = 0.39 to 0.66) for the six estimates taken by sex and test.
View this table:
[in this window]
[in a new window]
|
Table 3. Estimation of heritability calculated by ANOVA of the between-group ( B2) and within-group variance component ( W2) from 11 inbred strains (see Eq. 4), with the assumption that all variation is additive
|
|
 |
Discussion
|
---|
Much of the complexity of our existence emanates from operation of a sensorimotor network that can be modeled as having three primary components: 1) a central motor command that specifies the task, 2) a state change that implements the motor command, and 3) sensory feedback that provides information about the new state (29). The rotorod test we employed represents a simple-to-perform, yet complete, sensorimotor task that as a minimum engages the above three components. The substantial variation and relatively high estimates of heritability confirm our hypothesis that rotorod capacity is suitable for creation of low and high rat models of this trait by divergent artificial selection.
Defining the substrate suitable for the initiation of artificial selection is of interest because the models developed originally by selection can be much more informative for three primary reasons. First, the low and high line traits can often be made to differ substantially, which increases the signal measurements. Second, if the coefficient of inbreeding is kept low, then the major population complement of contrasting alleles causative of trait difference will be concentrated in the divergent lines. Third, selection across many generations interprets into inadvertent selection for lack of sensitivity to subtle differences in environment, such as room temperature and time of day of testing. This can be of large benefit because inbred strains that differ markedly for a trait, that did not originate from selection, often demonstrate wide trait variation in response to similar environments (4). A relative lack of sensitivity to minor differences in environment presumably explains the historical utility of two rat models of hypertension. Both the Dahl salt-sensitive (13) rats and the spontaneously hypertensive rats (30) were created initially by artificial selection.
The availability of N:NIH rats from the Animal Resource Center of the National Institutes of Health as a founder population is a major resource. This genetically segregating stock of rats originated from the intentional crossbreeding of eight disparate inbred strains (ACI, BN, BUF, F344, M520, MR, WKY, and WN) by Hansen and Spuhler in 1979 (9). Inbreeding drives all alleles toward homozygosity such that grossly deleterious recessive alleles were omitted with the creation of these inbred strains. The resultant outcross population has relatively wide genetic heterogeneity and is thus somewhat ideal as a starting population for selection. Based on R. A. Fishers 1930 Theorem of Natural Selection (8), traits peripherally associated with evolutionary fitness, such as morphology and complex physiology, demonstrate more additive genetic variance (i.e., h2) than traits essential to fitness because of less pressure from natural selection. Thus the relatively high h2 values for the rotorod test demonstrate that the N:NIH stock contain a diversity of allelic variants for this trait. As an extension, these observations support the argument that substrates challenged by the rotorod test are not closely associated with evolutionary fitness.
With artificial selection, it is not the trait itself that is selected upon, but a test or measure of a trait. It is axiomatic that many inextricably intertwined traits or "likely determinant phenotypes" such as coordination, balance, strength, subtle anatomic differences, fear, tractability, intelligence, and motivation will contribute to a rats capacity to perform on the rotorod (2). Traits of this complexity are termed multifactorial to emphasize their determination by multiple genetic and environmental factors. From the genetical perspective, traits can be divided into either Mendelian or quantitative. Mendelian traits are those for which a genetic difference at a single locus is sufficient to cause a difference in phenotypic expression of a given character. Quantitative traits do not manifest as discrete phenotypes in populations, but distribute with continuous variation. Continuous variation, such as that found for rotorod capacity, is the result of the variable presence and expression of many genes (i.e., polygenic) within a population as they interact with the environment. Overall, the outcome of a selection program on a quantitative trait is determined by 1) the summed influence of genes segregating for the trait in the founder population, and 2) naturally occurring mutations over generations producing new allelic variation for the trait.
Animal models first selectively bred for contrasting low and high measures of a trait, and then subsequently developed into inbred strains, are of substantial value. The major value of inbred strains emanates from their close genetic uniformity that facilitates genotyping, phenotyping, and the opportunity for multiple investigators to evaluate the same genetic substrate repeatedly; this type of uniformity cannot be approached in human studies.
Cenci and colleagues (3) recently summarized the relevance of rat models for evaluation of neurological deficits. As expected from knowledge of genome similarities within mammals (15), comparative studies disclose growing evidence for numerous similarities in sensorimotor function at all levels of organization in species as apparently different as rodents and primates (3). The first comprehensive summary of the Human Genome Project (26) provided evidence for 26,588 protein encoding transcripts for which
41% could not be classified and were termed proteins with unknown function. The ultimate creation of generalized robust genetic models of low and high sensorimotor capacity would be important tools for resolving variation in this trait at the genetic and molecular levels of organization. Identifying alleles that are causative of the difference between low and high sensorimotor capacity would provide fundamental information for creation of gene-based therapeutic intervention and be useful in defining function for unknown proteins. Comparative studies in well-defined models will play an important role in annotation (15).
 |
ACKNOWLEDGMENTS
|
---|
We acknowledge the contributions of Paul H. Brand to this work. We thank Krista Pettee, Jonathan Shields, and Mary Nelson for technical assistance, and we thank Marianne Jasper for the preparation of this manuscript.
This work was supported by National Heart, Lung, and Blood Institute Grant HL-64270.
 |
FOOTNOTES
|
---|
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).
Address for reprint requests and other correspondence: L. G. Koch, Physiology and Molecular Medicine, Medical College of Ohio, 3035 Arlington Ave., Toledo, OH 43614-5804 (E-mail: lkoch{at}mco.edu).
10.1152/physiolgenomics.00137.2002
 |
References
|
---|
- Auvray N, Caston J, Reber A, and Stelz T. Role of the cerebellum in the ontogenesis of the equilibrium behavior in the young rat: a behavioral study. Brain Res 505: 291301, 1989.[ISI][Medline]
- Biesiadecki BJ, Brand PH, Koch LG, Metting PJ, and Britton SL. Phenotypic variation in sensorimotor performance among eleven inbred rat strains. Am J Physiol Regul Integr Comp Physiol 276: R1383R1389, 1999.[Abstract/Free Full Text]
- Cenci MA, Whishaw IQ, and Schallert T. Animal models of neurological deficits: how relevant is the rat? Nat Rev Neurosci 3: 574579, 2002.[ISI][Medline]
- Crabbe JC, Wahlsten D, and Dudek BC. Genetics of mouse behavior: Interactions with laboratory environment. Science 284: 16701672, 1999.[Abstract/Free Full Text]
- Dunham NW and Miya TS. A note on a simple apparatus for detecting neurological deficit in rats and mice. J Am Pharm Assoc (Wash) 46: 208209, 1957.
- Falconer DS and Mackay TFC. Heritability. In: Introduction to Quantitative Genetics. Essex: Addison Wesley Longman, 1996, 160183.
- Falconer DS and Mackay TFC. Introduction to Quantitative Genetics. Essex: Addison Wesley Longman, 1996.
- Fisher RA. The Genetical Theory of Natural Selection. Oxford, UK: Clarendon, 1930.
- Hansen C and Spuhler K. Development of the National Institutes of Health genetically heterogeneous rat stock. Alcohol Clin Exp Res 8: 477479, 1984.[ISI][Medline]
- Hartl DL and Clark AG. Principles of Population Genetics. Sunderland: Sinauer, 1988.
- Hegmann JP and Possidente B. Estimating genetic correlations from inbred strains. Behav Genet 11: 103114, 1981.[ISI][Medline]
- Jones BJ and Roberts DJ. The quantitative measurement of motor inco-ordination in naive mice using an accelerating rotarod. J Pharm Pharmacol 20: 302304, 1968.[ISI][Medline]
- Knudsen K, Dahl LK, Thompson K, Iwai J, and Leith G. Effects of chronic salt ingestion: inheritance of hypertension in the rat. J Exp Med 132: 9761000, 1970.[ISI][Medline]
- Koch LG and Britton SL. Artificial selection for intrinsic aerobic endurance running capacity in rats. Physiol Genomics 5: 4552, 2001.[Abstract/Free Full Text]
- Lander ES et al. (International Human Genome Sequencing Consortium). Initial sequencing and analysis of the human genome. Nature 409: 860921, 2001.[ISI][Medline]
- Maes HHM, Beunen GP, Vlietinck RF, Neale MC, Thomis M, Vanden Eyne B, Lysens R, Simons J, Derom C, and Derom R. Inheritance of physical fitness in 10-yr-old twins and their parents. Med Sci Sports Exerc 28: 14791491, 1996.[ISI][Medline]
- Marisi DQ. Genetic and extragenetic variance in motor performance. Acta Genet Med Gemellol (Roma) 26: 197204, 1977.[Medline]
- Mousseau TA and Roff DA. Natural selection and the heritability of fitness components. Heredity 9: 181197, 1987.
- Murphy MP, Rick JT, Milgram NW, and Ivy GO. A simple and rapid test of sensorimotor function in the aged rat. Neurobiol Learn Mem 64: 181186, 1995.[ISI][Medline]
- Reed TE. Narrow-sense heritability estimates for nerve conduction velocity and residual latency in mice. Behav Genet 18: 595603, 1988.[ISI][Medline]
- Reed TE. Residual latency (delay at the neuromuscular junction): normative values and heritability in mice. Behav Genet 14: 209219, 1984.[ISI][Medline]
- Rijsdijk FV, Boomsma DI, and Vernon PA. Genetic analysis of peripheral nerve conduction velocity in twins. Behav Genet 25: 341348, 1995.[ISI][Medline]
- Sokal RR and Rohlf FJ. Biometry: The Principles and Practice of Statistics in Biological Research. San Francisco: Freeman, 1981.
- Swallow JG, Carter PA, and Garland T Jr. Artificial selection for increased wheel-running behavior in house mice. Behav Genet 28: 227237, 1998.[ISI][Medline]
- Vandenberg SG. The hereditary abilities study: hereditary components in a psychological test battery. Am J Hum Genet 14: 220237, 1962.[ISI]
- Venter JC et al. (Celera Genomics). The sequence of the human genome. Science 291: 13041351, 2001.[Abstract/Free Full Text]
- Williams LRT and Gross JB. Heritability of motor skill. Acta Genet Med Gemellol (Roma) 29: 127136, 1989.
- Williams LRT and Hearfield V. Heritability of a gross motor balance task. Res Q Exerc Sport 44: 109112, 1973.
- Wolpert DM and Ghahramani Z. Computational principles of movement neuroscience. Nat Neurosci 3: 12121217, 2000.[Medline]
- Yamori Y, Ooshima A, and Okamoto K. Genetic factors involved in spontaneous hypertension in rats an analysis of F2 segregate generation. Jpn Circ J 36: 561568, 1972.[Medline]