Things are better if bivariate or multivariate analyses are carried out. Knowing that two characters have a strong genetic correlation tells us that at least a part of the pathways leading from genotype to phenotype are shared and that a functional relationship may exist between the variables being studied . Indeed, if one peruses recent issues of the journal Behavior Genetics (where a large part of human quantitative-genetic research is published), it becomes
rapidly evident that bivariate and multivariate analyses find more have become the norm and that authors strive to uncover functional relationships between different behavioral phenotypes by investigating genetic correlations (see, e.g., ). In addition, human behavior geneticists are starting to follow the example of psychiatric geneticists and are attempting to localize genes involved in the regulation of behavior, with up till now equally mixed results, however (e.g., 34 and 35]). Gene localization, unfortunately, is remaining an elusive goal in human behavior genetics. Whereas classical linkage studies were (and PS-341 datasheet are) powerful tools to find genes for monogenic disorders, they have dismally failed to help us
identifying genes involved in either normal or pathological behavior. However, since the completion of the human genome project, Genome-Wide Association Studies (GWAS) are helping gene identification for polygenic disorders and have been particularly successful in non-mental disorders like cardiovascular disease (e.g., ). The first positive results are coming in for psychopathologies but huge sample sizes were needed before the first genomic risk loci could be identified . Recent efforts to identify genomic risk loci involved in schizophrenia have used total sample
sizes (including controls) of up to 150 000 subjects . In principle, of course, there is not really any reason to divide behavior genetics into human and animal studies. In practice, however, the ability to manipulate populations and to carry out directed breeding means that the field has advanced much farther in animal species. Other articles in this issue will present results obtained with fruit flies, worms, and fish. Here I concentrate on mouse studies. As in humans, animal behavior genetics used to be Methocarbamol heavily oriented towards quantitative genetics, using such designs as the classical Mendelian cross  or the diallel cross 40 and 41]. In addition, however, selection studies and comparisons between inbred strains often allowed research into the neural mechanisms underlying behavioral differences. Already in the late sixties and early seventies, the pharmacogenetic experiments of van Abeelen referred to above allowed him to conclude that C57/BL6J mice had a hippocampus that functions more effectively than that of DBA/2J mice (for a review, see [10•]), a conclusion confirmed many times since. Since then, many new tools have become available to study the genetics of behavior.