Approaches to handling incomplete data in family-based association testing.

The high throughput of data arising from the complete sequence of the human genome has left statistical geneticists with a rich and extensive information source. The wide availability of software and the increase in computing power has improved the possibilities to access and process such data. One problem is incompleteness of the data: unobserved or partially observed data points due to technical reasons or reasons associated with the patient's status or erroneous measurements of phenotype or genotype, to name a few. When not properly accounted for, these sources of incompleteness may seriously jeopardize the credibility of results from analyses. In this paper we provide some perspectives on the occurrence and analysis of different forms of incomplete data in family-based genetic association testing.

Allele dose analysis in recombinant inbred strains: a tool for multiple phenotype analysis with implications for quantitative trait loci mapping.

A considerable investment in genetic mapping is necessary to confirm linkages of quantitative trait loci for complex traits such as ethanol sensitivity, a significant predictor of alcoholism. Before embarking on such intensive mapping efforts in large intercrosses, we suggest an approach based on genetic marker data in recombinant strains that yield a rationale for selecting a battery of related phenotypes for confirmation studies of quantitative trait loci action. Using this approach with selected strains of mice, we retrospectively consider the relationship between ethanol sensitivity and neurotensin levels in several mammalian brain regions.

Genome-wide search for schizophrenia susceptibility loci: the NIMH Genetics Initiative and Millennium Consortium.

Schizophrenia has a complex pattern of inheritance, indicative of interactions among multiple genes and environmental factors. The detection and replication of specific susceptibility loci for such complex disorders are facilitated by the availability of large samples of affected sib pairs and their nuclear families, along with standardized assessment and systematic ascertainment procedures. The NIMH Genetics Initiative on Schizophrenia, a multisite collaborative study, was established as a national resource with a centralized clinical data base and cell repository. The Millennium Schizophrenia Consortium has completed a genome-wide scan to detect susceptibility loci for schizophrenia in 244 individuals from the nuclear families of 92 independent pairs of schizophrenic sibs ascertained by the NIMH Genetics Initiative. The 459 marker loci used in the scan were spaced at 10-cM intervals on average. Individuals of African descent were higher than those of European descent in their average heterozygosity (79% vs. 76%, P < .0001) and number of alleles per marker (9.2 vs. 8.4, P < .0001). Also, the allele frequencies of 73% of the marker loci differed significantly (P < .01) between individuals of European and African ancestry. However, regardless of ethnic background, this sample was largely comprised of schizophrenics with more than a decade of psychosis associated with pervasive social and occupational impairment.

NIMH Genetics Initiative Millenium Schizophrenia Consortium: linkage analysis of African-American pedigrees.

The NIMH Genetics Initiative is a multi-site collaborative study designed to create a national resource for genetic studies of complex neuropsychiatric disorders. Schizophrenia pedigrees have been collected at three sites: Washington University, Columbia University, and Harvard University. This article-one in a series that describes the results of a genome-wide scan with 459 short-tandem repeat (STR) markers for susceptibility loci in the NIMH Genetics Initiative schizophrenia sample-presents results for African-American pedigrees. The African-American sample comprises 30 nuclear families and 98 subjects. Seventy-nine of the family members were considered affected by virtue of having received a DSMIII-R diagnosis of schizophrenia (n = 71) or schizoaffective disorder, depressed (n = 8). The families contained a total of 42 independent sib pairs. While no region demonstrated evidence of significant linkage using the criteria suggested by Lander and Kruglyak, several regions, including chromosomes 6q16-6q24, 8pter-8q12, 9q32-9q34, and 15p13-15q12, showed evidence consistent with linkage (P = 0.01-0.05), providing independent support of findings reported in other studies. Moreover, the fact that different genetic loci were identified in this and in the European-American samples, lends credence to the notion that these genetic differences together with differences in environmental exposures may contribute to the reported differences in disease prevalence, severity, comorbidity, and course that has been observed in different racial groups in the United States and elsewhere.

Genome scan of European-American schizophrenia pedigrees: results of the NIMH Genetics Initiative and Millennium Consortium.

The Genetics Initiative of the National Institute of Mental Health (NIMH) was a multisite study that created a national repository of DNA from families informative for genetic linkage studies of schizophrenia, bipolar disorder, and Alzheimer's disease. The schizophrenia families were collected by three sites: Washington University, Harvard University, and Columbia University. This article, one in a series that describes the data collected for linkage analysis by the schizophrenia consortium, presents the results for the European-American sample. The European-American sample comprised 43 nuclear families and 146 subjects. Ninety-six of the family members were considered affected by virtue of having received a DSM-III-R diagnosis of schizophrenia (N = 82) or schizoaffective disorder, depressed (N = 14). The families contained a total of 50 independent sib-pairs. Using the significance threshold criteria suggested by Lander and Kruglyak [(1995): Nat Genet 241-247], no region showed statistically significant evidence for linkage; two markers on chromosome 10p showed statistical evidence suggestive of linkage using the criteria of Lander and Kruglyak [(1995): Nat Genet 241-247]: D10S1423 (nonparametric linkage (NPL) Z = 3.4, P = .0004) and its neighbor, D10S582 (NPL Z = 3.2, P = .0006).

Theoretical and empirical issues for marker-assisted breeding of congenic mouse strains.

Congenic breeding strategies are becoming increasingly important as a greater number of complex trait linkages are identified. Traditionally, the development of a congenic strain has been a time-consuming endeavour, requiring ten generations of backcrosses. The recent advent of a dense molecular genetic map of the mouse permits methods that can reduce the time needed for congenic-strain production by 18-24 months. We present a theoretical evaluation of marker-assisted congenic production and provide the empirical data that support it. We present this 'speed congenic' method in a user-friendly manner to encourage other investigators to pursue this or similar methods of congenic production.

Novel mutations and a mutational hotspot in the MODY3 gene.

Maturity-onset diabetes of the young 3 (MODY3) is a type of NIDDM caused by mutations in the transcription factor hepatocyte nuclear factor-1alpha (HNF-1alpha) located on chromosome 12q. We have identified four novel HNF-1alpha missense mutations in MODY3 families. In four additional and unrelated families, we observed an identical insertion mutation that had occurred in a polycytidine tract in exon 4. Among those families, one exhibited a de novo mutation at this location. We propose that instability of this sequence represents a general mutational mechanism in MODY3. We observed no HNF-1alpha mutations among 86 unrelated late-onset diabetic patients with relative insulin deficiency. Hence mutations in this gene appear to be most strongly associated with early-onset diabetes.

Common quantitative trait loci for alcohol-related behaviors and central nervous system neurotensin measures: hypnotic and hypothermic effects.

Genetic correlations were found between high-affinity neurotensin receptor (NTR(H)) densities and NT-immunoreactivity (NT-ir) levels in specific brain regions and sensitivity to hypnotic and hypothermic effects of ethanol in LSXSS recombinant inbred strains of mice. Simple sequence length polymorphisms were used to identify quantitative trait loci (QTL) influencing hypnotic and hypothermic sensitivity to ethanol, NTR(H) and low-affinity neurotensin receptor densities and NT-ir levels in LSXSS recombinant inbred strains. Common QTL for NTR(H) receptor densities, NT-ir levels and these ethanol actions were identified. One of the QTL (chromosome 2, 80 cM) for NTR(H) density and hypnotic sensitivity is linked to the NTR(H) gene, Ntsr. Also, QTL for NTR(H) density were found in common with confirmed QTL for hypnotic sensitivity on chromosomes 1 (43 cM), 11 (57 cM) and 15 (56 cM) and with an unconfirmed QTL on chromosome 3 (19 cM). Two common QTL for NT-ir levels, but not NTR(H) or low-affinity neurotensin receptor receptors, and ethanol-induced hypothermia were observed on chromosomes 4 (43 cM) and 6 (41 cM). Two common QTL for NT-ir levels and sleep time were identified on chromosomes 3 (19 cM) and 9 (55 cM). Common QTL indicate that genes regulating NT receptor and/or NT-ir expression may be the same as those regulating sensitivity to ethanol.

Confirmation of quantitative trait loci for ethanol sensitivity in long-sleep and short-sleep mice.

Initial insensitivity to alcohol is a strong predictor of human alcoholism, a widespread and heritable health problem. The Long Sleep and Short Sleep lines of mice were developed by genetic selection for high or low alcohol sensitivity. We have identified seven quantitative trait loci (QTLs) specifying differences in alcohol sensitivity using intercross progeny from these selected strains. These QTLs (Lorel-Lore7) together account for approximately 60% of the total genetic variance for this trait. This represents the first report of linkages for genes influencing alcohol action in any mammalian system using stringent, genome-wide mapping criteria.

Quantitative trait loci for ethanol sensitivity in the LS x SS recombinant inbred strains: interval mapping.

We are mapping the genes (quantitative trait loci or QTLs) that are responsible for individual differences in ethanol sensitivity, measured as the duration of loss of righting, reflex (LORR) and blood ethanol concentrations upon recovery of the righting reflex (BEC). The Long-Sleep (LS) and Short-Sleep (SS) selected lines of mice manifest an 18-fold difference in LORR and serve as a rodent model for ethanol sensitivity. The LS x SS recombinant inbred (RI) series, developed from LS and SS lines, are an important resource for QTL mapping of ethanol-related responses. The current report summarizes the initial QTL analysis of LORR and BEC in the LS x SS strains and compares the results of correlational analysis with an interval-mapping approach. The data provide strong evidence for QTLs that influence ethanol sensitivity on mouse chromosomes 1 and 2 and possible QTLs on chromosomes 1, 3, 4, 5, 6, 7, 12, 13, 16, and 18. These results are compared to those from an F2 cross which confirms QTLs on chromosomes 1, 2, 4, and 18.

Strain distribution patterns for genetic markers in the LSXSS recombinant-inbred series.

We present the strain distribution patterns (SDPs) of 118 SSLP markers and three pigmentation genes that have been characterized in 27 strains from the LSXSS RI series. This coarse map provides a resource for linkage studies of phenotypes that are heritable in the LSXSS RI series. The LSXSS recombinant inbred (RI) strains were derived from the Long-Sleep (LS) and Short-Sleep (SS) selected lines of mice that were selected for differential sensitivity to ethanol but are also differentially sensitive to a variety of other alcohols, barbiturates, sedative hypnotics, and general anesthetics. Since the parents were not inbred, two atypical factors are present in these SDPs. First, more than two alleles are frequently found in these RIs, and second, some alleles can be uniquely associated with one or the other parent while other alleles may be found in both parental lines. To validate the markers found in the parental line, we genotyped all parental mice from one generation of both the LS and SS lines, thus leading to a set of marker SDPs that are useful for further phenotypic association and identification of provisional QTLs.

Quantitative trait locus analyses of sleep-times induced by sedative-hypnotics in LSXSS recombinant inbred strains of mice.

The long-sleep (LS) and short-sleep (SS) selected lines of mice show highly significant differences in sleep-time for many sedative-hypnotic drugs, and the quantitative genetic nature of these differences has been well-established. Using an interval-mapping approach, quantitative trait locus (QTL) analyses of LSXSS recombinant inbred (RI) strains have been applied to sleep-time responses for various classes of sedative-hypnotic drugs: alcohols (ethanol, n-propanol, and n-butanol), the atypical anesthetic chloral hydrate, barbiturates (pentobarbital and secobarbital), and benzodiazepines (chlordiazepoxide and flurazepam). Several provisional QTLs were mapped to similar locations within and between drug classes, suggesting that some common loci are involved in sleep-times elicited by these drugs. Consistent with correlations of strain mean sleep-times between drugs tested in the LSXSS recombinant inbred strains, the number of provisional QTLs mapping to the locations of highest significance for ethanol decreases when the lipid solubility of a particular drug becomes less similar to that of ethanol. Provisional QTLs mapped for the benzodiazepines, however, revealed considerable overlap with those mapped for ethanol, although these drugs represented the most lipid-soluble category of sedative-hypnotics tested. Provisional QTLs for pentobarbital and secobarbital differed from most of those mapped for the alcohols, which supports the hypothesis that alcohols and barbiturates exert their effects mainly through different biological mechanisms in the LS and SS lines. Blood ethanol concentrations at regaining the righting reflex also mapped to several provisional QTLs corresponding to ethanol-induced sleep-times that support the contention that sleep-time is a reasonable index of the observed differences in central nervous system sensitivities to ethanol between LS and SS mice.

Use of repeated measures in an analysis of ethanol-induced loss of righting reflex in inbred long-sleep and short-sleep mice.

We present a repeated-measures analysis of ethanol-induced loss of righting reflex (LRR) in Inbred Long-Sleep (ILS) and Inbred Short-Sleep (ISS) strains of mice and their F1 and F2 cross progeny. Mice were administered a 4.1 g/kg intraperitoneal dose of ethanol at two times, 7-10 days apart. Repeatability is nonsignificant in ILS, ISS, and F1 mice, but is highly significant (0.47, p < 0.01) in the F2 mice. Mean LRR does not differ between trials 1 and 2, with the exception of the ISS strain in which the interaction of sex with LRR sensitization is significant. This two-trial method leads to increased accuracy of genotype assessment for pharmacological or behavioral traits where trial 1 does not influence the outcome of trial 2. The repeated-measures design facilitates novel analyses of the duration of LRR, and results suggest that most environmental variance for LRR is due to nonreplicable influences.

Ethanol-induced anesthesia in inbred strains of long-sleep and short-sleep mice: a genetic analysis of repeated measures using censored data.

We present a repeated-measures, genetic analysis of ethanol-induced anesthesia (sleep time) in mice from the inbred long-sleep (ILS) and short-sleep (ISS) strains of mice and their derived F1 and F2 generations. Mice (total N > 1300) were administered a 4.1 g/kg intraperitoneal dose of ethanol at two times, 7 to 10 days apart. A method suggested by Cohen was used to analyze the resulting data, which contained a known number of unmeasured observations. A genetic model that included both additive and epistatic parameters fitted the adjusted data well. A novel procedure for estimating heritability, based on a comparison of repeatability coefficients from the isogenic and genetically segregating populations, was also employed. Results of this analysis suggested that the heritability for individual measurements of sleep time is about .30, whereas the heritability of average scores is about .40. Results of this analysis also confirm that the ILS and ISS strains differ for genes at seven or more loci that influence sleep time. Since the ILS and ISS were derived by inbreeding the LS and SS lines, these results further suggest that almost all of the genetic variation present in the selected lines have been maintained in the ILS and ISS strains.

Initial characterization of STS markers in the LSXSS series of recombinant inbred strains.

We are mapping quantitative trait loci (QTLs) that influence ethanol-induced anesthesia (sleep time) in the Long-sleep (LS) and Short-sleep (SS) selected lines of mice. Fifty microsatellite-STS markers were initially screened for simple-sequence length polymorphisms between the LS and SS lines. Nineteen markers were polymorphic. Eleven markers unequivocally differentiated the LS and SS lines and were used to establish strain distribution patterns for the LSXSS series of recombinant inbred strains. Five markers each accounted for at least 5% of sleep-time genetic variance among the RI strains. Linkage of provisional QTLs detected among RIs will be confirmed or disproved in a large F2 population. This ongoing QTL-mapping project eventually will result in a strain distribution pattern for the LSXSS RI series with an average marker spacing of 5 centimorgans.

A multivariate analysis of repeated measures: linkage of the albinism gene (Tyr) to a QTL influencing ethanol-induced anesthesia in laboratory mice.

We propose the use of multivariate analysis of variance in order to address linkage relationships between a genetic marker and a phenotype that is specified by multiple variables. Moreover, enhanced power of a linkage model that simultaneously accounts for environmental factors is demonstrated in the present study. This strategy was used to establish linkage of a quantitative trait locus influencing ethanol-induced anesthesia to the albinism gene (Tyr) in mouse. This approach, which can be implemented immediately with most statistical packages, provides the extreme flexibility of multivariate analysis of variance for analyzing complex questions of linkage regarding either animal or human research paradigms.

Mapping quantitative trait loci for ethanol-induced anesthesia in LSxSS recombinant inbred and F2 mice: methodology and results.

QTL (quantitative trait locus) mapping is an exciting new technique currently being applied to many quantitative traits. We review the animal models and molecular genetic techniques of QTL mapping with particular reference to ethanol-induced anesthesia (sleep time). Sleep time is an easily obtained measure of initial sensitivity to ethanol. We are using three-stage strategy to map QTLs causing the difference in sleep time between selected strains of mice. This project has been expedited by the use of polymorphic genetic markers (simple sequence repeats) which are amplifiable by the polymerase chain reaction. Over 115 strain distribution patterns have been characterized in the 27 LSxSS recombinant inbred strains. Provisional QTLs have been identified in the RI strains and will be confirmed or rejected in an F2 population of 1063 mice. A new statistical technique, interval mapping, allows the use of all polymorphic markers simultaneously to locate QTLs.

Mapping quantitative trait loci for behavioral traits in the mouse.

After many years of studying various behavioral characters in the mouse, it is clear that most are heritable and are specified by complexes of genes or quantitative trait loci (QTLs). In order to attain a more complete understanding of the genetic causes of individual differences in behavior, the mechanism of action of these QTLs must be elucidated. The most straightforward approach to determining the mechanism of action of a particular QTL is to identify and molecularly clone the gene; this can be done by positional cloning, which depends on precise knowledge of the genetic map position. As the genetic data base for the mouse genome continues to develop, such strategies will become increasingly easy to perform. Here we suggest a multistage strategy for QTL mapping using recombinant-inbred strains of mice: (1) characterize genomic DNA from parental strains originally used to generate the RI strains; (2) characterize the RI strains for a quantitative character and for DNA markers that differ in the parental strains; and (3) assess the quantitative character in F2 mice derived from crosses between the parental strains, then determine the genotypes of extreme F2 mice for markers that account for at least 5% of the additive genetic variance. Data from these F2 crosses can be used to test hypotheses from the analysis of RI strains, i.e., that a QTL maps to a particular region. Using data from the mouse genome data base, this strategy should allow the molecular identification of the gene based on a candidate-gene approach. We illustrate the approach with examples from our work in mapping QTLs specifying neural sensitivity to the anesthetic effects of ethanol.