Haplotype Sharing Refines the Location of an Imprinted Quantitative Trait Locus With Major Effect on Muscle Mass to a 250-kb Chromosome Segment Containing the Porcine IGF2 Gene

Corresponding author: Michel Georges, Faculty of Veterinary Medicine, University of Liège (B43), 20 Bd de Colonster, 4000-Liège, Belgium., michel.georges{at}ulg.ac.be (E-mail)

Communicating editor: C. HALEY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We herein describe the fine mapping of an imprinted QTL with major effect on muscle mass that was previously assigned to distal SSC2p in the pig. The proposed approach exploits linkage disequilibrium in combination with QTL genotyping by marker-assisted segregation analysis. By identifying a haplotype shared by all "Q" chromosomes, we map the QTL to an 250-kb chromosome segment containing INS and IGF2 as the only known paternally expressed genes. This considerably reinforces the candidacy of these genes, justifying their detailed analysis.

QUANTITATIVE trait loci (QTL) mapping has become a preferred approach toward the molecular dissection of quantitative traits, whether of fundamental, medical, or agronomic importance. A multitude of chromosomal locations predicted to harbor genes influencing traits of interest have been identified using this strategy (e.g., ANDERSSON 2001 ; FLINT and MOTT 2001 ; MACKAY 2001 ; MAURICIO 2001 ). In most cases, however, the mapping resolution is in the order of the tens of centimorgans, which is insufficient for positional cloning of the underlying genes. High-resolution mapping of QTL therefore remains one of the major challenges in the genetic analysis of complex traits.

Three factors limit the achievable mapping resolution: marker density, crossover density, and the ability to deduce QTL genotype from phenotype. Increasing marker density may still be time-consuming in many organisms but is conceptually the simplest bottleneck to resolve. Two options are available to increase the local crossover density: breed recombinants de novo or exploit historical recombination events; i.e., use linkage disequilibrium (LD). The former approach is generally used with model organisms that have a short generation interval (e.g., DARVASI 1998 ), while the latter is the only practical alternative when working with human or large livestock species. Optimal use of LD to fine map QTL in outbred populations is presently an area of very active research (e.g., ARDLIE et al. 2002 ). The ability to deduce QTL genotype from phenotype can be improved by using "clones" (e.g., recombinant inbred lines; DARVASI 1998 ), by means of progeny testing (e.g., GEORGES et al. 1995 ), or by marker-assisted segregation analysis (e.g., RIQUET et al. 1999 ).

Recently a QTL with major effect on muscle mass and fat deposition was mapped to the distal end of chromosome arm SSC2p in the pig (JEON et al. 1999 ; NEZER et al. 1999 ). The most likely position of the QTL was shown to coincide with a chromosome region that is orthologous to HSA11p15 in the human, which is known to harbor an imprinted domain. QTL analyses performed with imprinting models strongly suggested that the underlying gene was indeed imprinted and expressed only from the paternal allele (JEON et al. 1999 ; NEZER et al. 1999 ). The human 11p15 imprinted domain is known to contain at least 9 imprinted transcripts. Three of these are paternally expressed: LIT-1 (KVLQT1-AS), IGF2, and IGF2-AS (e.g., REIK and WALTER 2001 ). Fifteen imprinted transcripts are known to map to the orthologous domain on distal mouse chromosome MMU7, of which 4 are paternally expressed: Lit-1 (Kvlqt1-as), Ins2, Igf2, and Igf2-as (e.g., http://www.mgu.har.mrc.ac.uk/imprinting/imprinting.html; ONYANGO et al. 2000 ). Because of its known function in myogenesis (FLORINI et al. 1996 ), IGF2 stood out as a prime positional candidate. However, no sequence variations that could account for the observed QTL effect were found in the coding parts of the porcine IGF2 gene (JEON et al. 1999 ; NEZER et al. 1999 ).

To refine the map position of this QTL and to verify whether its position remained compatible with a direct role of the IGF2 gene, we applied an approach targeting the three factors limiting the mapping resolution of QTL: (i) we increased the marker density of the chromosome region of interest; (ii) we determined the QTL genotype of a number of individuals by marker-assisted segregation analysis; and (iii) we applied a LD-based haplotype-sharing approach to determine the most likely position of the QTL. This approach is analogous to the one that was previously applied by RIQUET et al. 1999 to refine the map position of a QTL influencing milk production in dairy cattle. It makes the assumption that the observed QTL effect is due to the segregation of a QTL allele with major substitution effect ("Q") that appeared by mutation or migration g generations ago and swept through the populations as a result of artificial selection. As a consequence, at the present generation, n chromosomes carrying the Q allele are expected to share a haplotype of size 2/ng (in morgans) containing the QTL (DUNNER et al. 1997 ).

By doing so we have identified a shared haplotype spanning 250 kb that is predicted to contain the quantitative trait nucleotide (QTN; MACKAY 2001 ). The corresponding chromosome segment contains INS and IGF2 as the only known paternally expressed genes. This considerably enforces the candidacy of these two genes and demonstrates that LD can be exploited to map QTL in outbred populations to chromosome intervals containing no more than a handful of genes.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Pedigree material and phenotypic data:
The pedigree material used for this work was composed of a subset of a previously described Piétrain x Large White F₂ pedigree (HANSET et al. 1995 ; NEZER et al. 2002 ), as well as a series of paternal half-sib pedigrees sampled in commercial lines derived from the Piétrain and Large White breeds (N. BUYS, personal communication). In the F₂ animals, percentage of lean meat was measured as "percentage of lean cuts" as previously described (HANSET et al. 1995 ), while in the commercial lines percentage of lean meat was measured as "Piglog" corresponding to (63.6882 - 0.4465 a - 0.5096 b + 0.1281 c), where a is millimeters of backfat measured between the third and fourth lumbar vertebra at 7 cm from the spine, b is millimeters of backfat measured between the third and fourth last rib at 7 cm from the spine, and c is millimeters of loin thickness, measured at same position as b.

Marker-assisted segregation analysis:
The QTL genotype of each sire was determined from a Z-score, corresponding to the log₁₀ of the likelihood ratio L_H1/L_H0, where L_H1 corresponds to the likelihood of the pedigree data assuming that the boar is of Qq genotype, and L_H0 corresponds to the likelihood of the pedigree data assuming that the boar is of QQ or qq genotype. The corresponding likelihoods were computed as

In this n is the number of informative offspring having inherited the "left" homolog from their sire, m is the number of informative offspring having inherited the "right" homolog from their sire, y_i(j) is the phenotype of offspring i (j), is the average phenotype of the corresponding pedigree computed over all (informative and noninformative) offspring, is the residual standard deviation maximizing L, and a is the Q to q allele substitution effect. |a| was set at zero when computing L_H0 and at 2.0% when computing L_H1 (NEZER et al. 1999 ). The sign of a (+ or -) was selected to maximize L_H1. Boars were considered to be Qq when Z > 2, QQ or qq when Z < -2, and of undetermined genotype if 2 > Z > -2.

Linkage disequilibrium analysis:
Probabilities for two chromosomes to be identical by descent (IBD) at a given map position conditional on flanking marker data were computed according to MEUWISSEN and GODDARD 2001 . The effective population size (N_e) was set at 200 on the basis of estimates of N_e determined from LD data (N. HARMEGNIES, unpublished observations) and the number of generations to the base population at 20. A multipoint test for association was performed using the DISMULT program described in TERWILLIGER 1995 .

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

QTL genotyping by marker-assisted segregation analysis:
We genotyped a series of paternal half-sib families, counting at least 20 offspring for two microsatellite markers located on the distal end of chromosome arm SSC2p: SWR2516 and SWC9 (JEON et al. 1999 ; NEZER et al. 1999 ). These families originated either from a previously described Piétrain x Large White F₂ pedigree (NEZER et al. 2002 ) or from two composite pig lines derived from Large White and Piétrain founder animals (N. BUYS, personal communication).

The pedigrees from sires that were heterozygous for one or both of these markers were kept for further analysis. Twenty such pedigrees could be identified for a total of 941 animals. Offspring were sorted in three classes on the basis of their marker genotype: "L" (left homolog inherited from the sire), "R" (right homolog inherited from the sire), or "?" (not informative or recombinant in the SWR2516-SWC9 interval).

Offspring were slaughtered at a constant weight of 105 kg, and a series of phenotypes were collected on the carcasses, including percentage of lean meat, measured as either percentage of lean cuts (experimental cross) or Piglog (composite lines; see MATERIALS AND METHODS).

We then computed the likelihood of each sire family under two hypotheses: H₀, postulating that the corresponding boar was homozygous at the QTL, and H₁ postulating that the boar was heterozygous at the QTL. Assuming a biallelic QTL, H₀ corresponds to QTL genotypes QQ or qq, and H₁ to genotype Qq. Likelihoods were computed using percentage of lean meat as phenotype (as the effect of the QTL was shown to be most pronounced on this trait in previous analyses) and assuming a Q to q allele substitution effect of 2.0% (NEZER et al. 1999 ). If the odds in favor of one of the hypotheses were superior or equal to 100:1, the most likely hypothesis was considered to be true (see MATERIALS AND METHODS).

Using these rules, we could determine the QTL genotype for 14 of the 20 boars. Seven of these proved to be heterozygous Qq and the other 7 to be homozygous and thus of either QQ or qq genotype (Fig 1).

View larger version (32K): In this window In a new window Download PPT slide   Figure 1. QTL genotyping by marker-assisted segregation analysis. The graphs show, for 14 paternal half-sib pedigrees (P1, P2, ... , P14), the phenotypic mean ±2 standard errors of the offspring sorted in two groups according to the homolog inherited from the sire. The number of offspring in each group is given above and under the error bars, respectively. The top graph corresponds to the boars that were shown to be heterozygous Qq for the QTL and the bottom graph to the boars that were shown to be homozygous at the QTL. Pedigrees for which the percentage of lean meat was measured as percentage of lean cuts (NEZER et al. 2002 ) are marked by L (left axis), those for which "Piglog" was used (see MATERIALS AND METHODS) are marked by R (right axis). The graph reports a Z-score for each pedigree, i.e., the log10 of the H1/H0 likelihood ratio where H1 assumes that the boar is heterozygous Qq for the QTL, while H0 assumes that the boar is homozygous QQ or qq. Q alleles associated with a positive allele substitution effect on percentage of lean meat are marked by a diamond, q alleles by a circle. The number within the symbols differentiates the Q and q alleles according to the associated marker haplotype (see RESULTS and Fig 2).

Constructing a physical and genetic map of the porcine ortholog of the human 11p15 imprinted domain:
We developed porcine sequence-tagged sites (STS) across the orthologous region of the human 11p15 imprinted domain. The SWC9 marker was known from previous studies to correspond to a (CA)_n microsatellite located in the 3' untranslated region (UTR) of the porcine IGF2 gene (JEON et al. 1999 ; NEZER et al. 1999 ; AMARGER et al. 2002 ). Fourteen novel STS were developed in genes [TSSC5, KVLQT1 (3x), CD81, TH (2x), INS (3x), IGF2 (3x), and H19] and 5 in intergenic regions [IG(IGF2-H19), IG(H19-RL23mrp) (4x)]. The corresponding primer sequences were derived from the porcine genomic sequence when available (AMARGER et al. 2002 ) or from porcine expressed sequence tags (EST) that were identified by BLAST searches using the human orthologs as query sequences (Table 1).

We screened a porcine bacterial artificial chromosome (BAC) library (FAHRENKRUG et al. 2001 ) by filter hybridization, using (i) human cDNA clones corresponding to genes know to map to 11p15, as well as (ii) some of the 19 previously described porcine STS. Six of the identified BACs were shown by PCR to contain at least one of the porcine STS available in the region and were kept for further analysis (BACs 466F9, 267M13, 389B2, 370C17, and 453A14). To this set we added two BACs that were previously shown to span the TH-H19 region (INRA370 and INRA253; AMARGER et al. 2002 ) and one that was isolated from the same library using a KVLQT1 probe (INRA956B11; E. GIUFFRA, unpublished results). Three additional STS were developed from BAC end sequences (389B2T7, 370C17T7, and INRA370SP6) and 4 from randomly selected subclones of BAC 370C17 (370SNP1, 370SNP2, 370SNP6, and 370SNP15; Table 1). INRA370SP6 and 370SNP2 revealed highly significant BLAST hits in the vicinity of the ASCL2 gene (expected values 10^-7) and TSSC6 gene (expected values 10^-18), respectively, providing additional anchor points between the human and porcine sequence. Using STS content mapping, we assembled the BAC contig shown in Fig 2. It is compatible with the overall conservation of gene order between human and pigs in this chromosome region and indicates that the gap remaining in the human sequence between the INS and ASCL2 genes may not be >55 kb.

View larger version (56K): In this window In a new window Download PPT slide   Figure 2. (A) Schematic representation of the human 11p15 imprinted domain according to ONYANGO et al. 2000 . Biallelically expressed genes are shown as lightly shaded cylinders, imprinted genes as darkly shaded cylinders. Maternally expressed genes are italicized and paternally expressed genes are underlined. (B) BAC contig spanning the porcine ortholog of the 11p15 imprinted domain, assembled by STS content mapping. The length of the horizontal bars does not reflect the actual physical size of the corresponding BACs. It is assumed in this map that the gene order with respect to telomere is conserved in man and pig (http://www.ensembl.org/homo_sapiens/). (C) Marker haplotypes of the five Q chromosomes (diamonds) and six q chromosomes (circles). Closely linked SNPs (<5 kb) or adjacent SNPs that could not be ordered were merged into polyallelic multisite haplotypes (cf. Table 2). The darkened chromosome segments correspond to the haplotype shared by all Q chromosomes and are therefore assumed to contain the QTL.

All available STS were then amplified from genomic DNA of the 14 QTL-genotyped boars (see above) and cycle sequenced to identify DNA sequence polymorphisms. We identified a total of 51 single-nucleotide polymorphisms (SNPs): 2 in TSSC5, 15 in KVLQT1, 3 in 389B2-T7, 1 in 370SNP15, 1 in 370SNP6, 3 in 370SNP1, 3 in 370SNP2, 4 in TH, 7 in INS, 4 in IGF2, 1 in IG(IGF2-H19), 3 in H19, and 4 in IG(H19-RL23MRP) (Table 1).

Three microsatellites were added to this marker list: one (KVLQT1(SSR) isolated from BAC INRA956B11 and two (PULGE1 and PULGE3) isolated from BAC INRA370 (Table 1).

Assembling pools of Q- vs. q-bearing chromosomes:
To reconstruct the marker linkage phase of the 14 QTL-genotyped sires, we selected—for each boar—offspring that were homozygous for the alternate paternal SWR2516-SWC9 haplotypes. These were genotyped for all SNPs and microsatellites available in the region, and from these genotypes we manually determined the linkage phase of the boars.

For six of the seven boars, shown by marker-assisted segregation analysis to be of Qq genotype (Fig 1), the Q chromosomes associated with an increase in percentage of lean meat proved to be identical by state (IBS) over their entire length. This haplotype was therefore referred to as Q¹. The haplotype corresponding to the seventh Q chromosome (P7 in Fig 2) was different and was referred to as Q². For three of these sires, the haplotypes associated with a decrease in percentage of lean meat proved to be completely IBS as well and were thus referred to as q¹. The other four q chromosomes carried distinct haplotypes and were referred to as q², q³, q⁴, and q⁵ (Fig 2).

The first boar that proved to be homozygous for the QTL by marker-assisted segregation analysis (P8) carried the q⁴ haplotype on one of its chromosomes. Its other haplotype therefore had to be of q genotype as well and was referred to as q⁶.

Boar P9 appeared to be heterozygous Q¹/Q². Boars P10 and P11 carried the Q¹ haplotype shared by six of the Qq boars. As a consequence, the other chromosomes of boars P10 and P11, which were IBS as well, were placed in the Q pool and referred to as Q³. Homozygous boar P12 carried haplotype Q². As a consequence, its homolog was referred to as Q⁴. Following the same recursive procedure, boars P13 and P14 were identified as being respectively Q³Q⁴ and Q²Q⁵.

The marker genotypes of the resulting five Q and five q chromosomes are shown in Fig 2. In this figure, closely linked (<5 kb) SNPs or SNPs that could not be ordered (370SNP15 with 370SNP6 and 370SNP1 with 370SNP2) were merged into a series of polyallelic multisite haplotypes. The correspondence between SNP genotype and haplotype number is given in Table 2.

All Q chromosomes share an 250-kb common haplotype encompassing the INS and IGF2 genes:
Visual examination of the Q and q pools immediately reveals that all five chromosomes in the Q pool share an IBS haplotype spanning the 370SNP1/2-IGF2 interval as predicted (Fig 2). Further examination of the Q chromosomes indicates that on the proximal side of the 370SNP1/2-IGF2 interval, the Q², Q³, and Q⁴ haplotypes are all identical, while the Q¹ and Q⁵ haplotypes are best explained by single-recombination events having occurred, respectively, in the 370SNP6/15-370SNP1/2 (Q¹) and KVLQT1(I7-389B27)(Q⁵) interval. On the distal side of the 370SNP1/2-IGF2 interval, three of the five Q chromosomes (Q¹, Q³, Q⁴) carry the same haplotype in the SWC9-IG(H19-RL23mrp) interval, while the two remaining ones (Q², Q⁵) are sharing a completely distinct one. Again, this is best explained by assuming a single ancestral recombination event just upstream of the SWC9 microsatellite marker. These observations therefore strongly suggest that the hypothesized Q allele causing an increase in percentage of lean meat appeared by mutation or migration on a founder chromosome carrying the haplotype highlighted in Fig 2 and that the QTL is consequently located in the 370SNP6/15-SWC9 interval.

No such shared haplotype could be identified in the q pool. As expected under our model, the q pool exhibited a higher level of genetic diversity. The q-bearing chromosomes would indeed be older, having had ample opportunity to recombine, thereby increasing haplotype diversity. This can be quantified more accurately by computing the average pairwise probability for Q and q chromosomes to be IBD conditional on flanking marker data, using the coalescent model developed by MEUWISSEN and GODDARD 2001 . As shown in Fig 3, the average pairwise IBD probability among the five Q chromosomes is superior to 0.25 over the entire KVLQT1(SSR)-IG(H19-RL23mrp) interval and exceeds 0.60 in the 370SNP1/2-IGF2 interval. For the q chromosomes, the equivalent parameter is inferior to 0.20 over the entire studied region.

View larger version (31K): In this window In a new window Download PPT slide   Figure 3. Average probability for two Q () and two q () chromosomes to be identical by descent at a given map position conditional on flanking marker data [P(IBD|MG)], along the chromosome segment encompassing the p57-H19 imprinted domain, computed according to MEUWISSEN and GODDARD 2001 . The positions of the markers defined according to Table 1 and Table 2 are shown by the vertical dotted lines.

One could argue that the probability of identifying a shared haplotype among five chromosomes by chance alone is high and does not support the location of the QTL within this region. To more quantitatively estimate the significance of the haplotype sharing among Q chromosomes, accounting for the distance between adjacent markers as well as allelic frequencies, we performed a multipoint LD analysis using the DISMULT program (TERWILLIGER 1995 ). To test the significance of the haplotype sharing observed among the five Q chromosomes, we performed the same DISMULT analysis on all 462 possible combinations of the 11 chromosomes taken 5 at a time. For each of these analyses we stored the highest likelihood obtained anywhere along the analyzed chromosome segment. The likelihood obtained using the five real Q chromosomes at the position of marker PULGE3 was the highest one obtained across all chromosome permutations (data not shown), clearly indicating that the observed haplotype sharing is very unlikely to be fortuitous.

At present, our best estimate of the size of the 370SNP6/15-SWC9 interval is of the order of 250 kb. From the knowledge of the orthologous region in the human, it is predicted (ONYANGO et al. 2000 ) to contain a maximum of eight biallelically expressed genes (TSSC9, TRPCL, TSSC4, CD81, TSSC10, RPL26, TSSC6, and TH), two imprinted genes expressed from the maternal allele (KVLQT1 and ASCL2), and two imprinted genes expressed from the paternal allele (INS and IGF2; Fig 2).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

When we previously demonstrated that only the paternal SSC2 QTL allele influenced muscle mass and that the most likely QTL position coincided with IGF2, this gene stood out as a prime positional candidate (JEON et al. 1999 ; NEZER et al. 1999 ). On the basis of the initial linkage analysis, however, the confidence interval for the QTL covered 4 cM, which were bound to contain a multitude of genes other than IGF2. It was therefore critical to refine the map position of the QTL, which we set out to do by exploiting both LD and marker-assisted segregation analysis. Because of the observed parent-of-origin effect, we focused our analysis on a chromosome region that is the ortholog of the human 11p15 imprinted domain. We herein provide strong evidence that the QTL indeed maps to the p57-H19 imprinted gene cluster and within this region to a chromosome segment of 250 kb containing INS and IGF2 as the only known paternally expressed genes. The proximal breakpoint of the interval indeed coincides with the SWC9 microsatellite marker known to be located in the 3' UTR of IGF2. These findings therefore considerably strengthen the candidacy of IGF2 in particular and justify its detailed analysis.

The fact that we succeeded in refining the map position of this QTL down to the subcentimorgan level supports its simple molecular architecture. Together with recent successes in positional cloning and identification of the mutations that underlie the major gene and QTL (e.g., GROBET et al. 1997 ; MILAN et al. 2000 ; BLOTT et al. 2002; FREKING et al. 2002 ; GRISART et al. 2002 ; SMIT et al. 2003 ), this clearly indicates that at least part of the genetic variation of production traits in livestock is due to single mutations with large effects on the traits of interest.

The success of haplotype-sharing approaches in fine mapping QTL in livestock also suggests that QTL may be mapped in these populations by virtue of the haplotype signature resulting from intense selection on Q alleles, i.e., haplotypes of unusual length given their population frequency. The feasibility of this approach has recently been demonstrated in human populations for two loci (G6PD and CD40 ligand) known to be involved in resistance to malaria (SABETI et al. 2002 ). Preliminary analyses suggest that similar haplotype signatures characterize the QTL described in this work as well as a positionally cloned QTL influencing milk yield and composition on bovine chromosome 14 (GRISART et al. 2002 ). QTL undergoing selection could thus be identified in livestock in the absence of phenotypic data.

	ACKNOWLEDGMENTS

We are grateful to Frédéric Farnir and Danny Wagenaar for their help with the ML and LD analysis. This work was supported by a grant from the Belgian "Ministère des Classes Moyennes et de l'Agriculture" and Gentec.

Manuscript received November 28, 2002; Accepted for publication May 13, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AMARGER, V., M. NGUYEN, A. S. VAN LAERE, C. NEZER, and M. GEORGES et al., 2002  Comparative sequence analysis of the INS-IGF2-H19 gene cluster in pigs. Mamm. Genome 13:388-398.[Medline]

ANDERSSON, L., 2001  Genetic dissection of phenotypic diversity in farm animals. Nat. Rev. Genet. 2:130-138.[Medline]

ARDLIE, K. G., L. KRUGLYAK, and M. SEIELSTAD, 2002  Patterns of linkage disequilibrium in the human genome. Nat. Rev. Genet. 3:299-309.[Medline]

BLOTT, S., J.-J. KIM, S. MOISIO, A. SCHMIDT-KÜNTZEL, and A. CORNET et al., 2003  Molecular dissection of a quantitative trait locus: a phenylalanine to tyrosine substitution in the transmembrane domain of the bovine growth hormone receptor is associated with a major effect on milk yield and composition. Genetics 163:253-266.[Abstract/Free Full Text]

DARVASI, A., 1998  Experimental strategies for the genetic dissection of complex traits in animal models. Nat. Genet. 18:19-24.[Medline]

DUNNER, S., C. CHARLIER, F. FARNIR, B. BROUWERS, and J. CANON et al., 1997  Towards interbreed IBD fine mapping of the mh locus: double-muscling in the Asturiana de los Valles breed involves the same locus as in the Belgian Blue cattle breed. Mamm. Genome 8:430-435.[Medline]

FAHRENKRUG, S. C., G. A. ROHRER, B. A. FREKING, T. P. SMITH, and K. OSOEGAWA et al., 2001  A porcine BAC library with tenfold genome coverage: a resource for physical and genetic map integration. Mamm. Genome 12:472-474.[Medline]

FLINT, J. and R. MOTT, 2001  Finding the molecular basis of quantitative traits: successes and pitfalls. Nat. Rev. Genet. 2:437-445.[Medline]

FLORINI, J. R., D. Z. EWTON, and S. A. COOLICAN, 1996  Growth hormone and the insulin-like growth factor system in myogenesis. Endocr. Rev. 17:481-517.[Medline]

FREKING, B. A., S. K. MURPHY, A. A. WYLIE, S. J. RHODES, and J. W. KEELE et al., 2002  Identification of the single base change causing the callipyge muscle hypertrophy phenotype, the only known example of polar overdominance in mammals. Genome Res. 12:1496-1506.[Abstract/Free Full Text]

GEORGES, M., D. NIELSEN, M. MACKINNON, A. MISHRA, and R. OKIMOTO et al., 1995  Mapping quantitative trait loci controlling milk production by exploiting progeny testing. Genetics 139:907-920.[Abstract]

GRISART, B., W. COPPIETERS, F. FARNIR, L. KARIM, and C. FORD et al., 2002  Positional candidate cloning of a QTL in dairy cattle: identification of a missense mutation in the bovine DGAT gene with major effect on milk yield and composition. Genome Res. 12:222-231.[Abstract/Free Full Text]

GROBET, L., L. J. ROYO MARTIN, D. PONCELET, D. PIROTTIN, and B. BROUWERS et al., 1997  A deletion in the myostatin gene causes double-muscling in cattle. Nat. Genet. 17:71-74.[Medline]

HANSET, R., C. DASNOIS, S. SCALAIS, C. MICHAUX, and L. GROBET, 1995  Effets de l'introgression dans le génome Piétrain de l'allèle normal au locus de sensibilité à l'halothane. Genet. Sel. Evol. 27:77-88.

JEON, J. T., O. CARLBORG, A. TÖRNSTEN, E. GIUFFRA, and V. AMARGER et al., 1999  A paternally expressed QTL affecting skeletal and cardiac muscle mass in pigs maps to the IGF2 locus. Nat. Genet. 21:157-158.[Medline]

MACKAY, T. F. C., 2001  Quantitative trait loci in Drosophila. Nat. Rev. Genet. 2:11-20.[Medline]

MAURICIO, R., 2001  Mapping quantitative trait loci in plants: uses and caveats for evolutionary biology. Nat. Rev. Genet. 2:370-381.[Medline]

MEUWISSEN, T. H. and M. E. GODDARD, 2001  Prediction of identity by descent probabilities from marker-haplotypes. Genet. Sel. Evol. 33:605-634.[Medline]

MILAN, D., J. T. JEON, C. LOOFT, V. AMARGER, and A. ROBIC et al., 2000  A mutation in PRKAG3 associated with excess glycogen content in pig skeletal muscle. Science 288:1248-1251.[Abstract/Free Full Text]

NEZER, C., L. MOREAU, B. BROUWERS, W. COPPIETERS, and J. DETILLEUX et al., 1999  An imprinted QTL with major effect on muscle mass and fat deposition maps to the IGF2 locus in pigs. Nat. Genet. 21:155-156.[Medline]

NEZER, C., L. MOREAU, D. WAGENAAR, and M. GEORGES, 2002  Results of a whole genome scan targeting QTL for growth and carcass characteristics in a Piétrain x Large White intercross. Genet. Sel. Evol. 34:371-387.[Medline]

ONYANGO, P., W. MILLER, J. LEHOCZKY, C. T. LEUNG, and B. BIRREN et al., 2000  Sequence and comparative analysis of the mouse 1-megabase region orthologous to the human 11p15 imprinted domain. Genome Res. 10:1697-1710.[Abstract/Free Full Text]

REIK, W. and J. WALTER, 2001  Genomic imprinting: parental influence on the genome. Nat. Rev. Genet. 2:21-32.[Medline]

RIQUET, J., W. COPPIETERS, N. CAMBISANO, J.-J. ARRANZ, and P. BERZI et al., 1999  Identity-by-descent fine-mapping of QTL in outbred populations: application to milk production in dairy cattle. Proc. Natl. Acad. Sci. USA 96:9252-9257.[Abstract/Free Full Text]

SABETI, P. C., D. E. REICH, J. M. HIGGINS, H. Z. LEVINE, and D. J. RICHTER et al., 2002  Detecting recent positive selection in the human genome from haplotype structure. Nature 419:832-837.[Medline]

SMIT, M., K. SEGERS, L. G. CARRASCOSA, T. SHAY, and F. BARALDI et al., 2003  Mosaicism of solid gold supports the causality of a noncoding A-to-G transition in the determinism of the callipyge phenotype. Genetics 163:453-456.[Abstract/Free Full Text]

TERWILLIGER, J. D., 1995  A powerful likelihood method for the analysis of linkage disequilibrium between trait loci and one or more polymorphic marker loci. Am. J. Hum. Genet. 56:777-787.[Medline]

This article has been cited by other articles:

				X. Liu, H. Zhang, H. Li, N. Li, Y. Zhang, Q. Zhang, S. Wang, Q. Wang, and H. Wang Fine-Mapping Quantitative Trait Loci for Body Weight and Abdominal Fat Traits: Effects of Marker Density and Sample Size Poult. Sci., July 1, 2008; 87(7): 1314 - 1319. [Abstract] [Full Text] [PDF]

				T. Druet, S. Fritz, M. Boussaha, S. Ben-Jemaa, F. Guillaume, D. Derbala, D. Zelenika, D. Lechner, C. Charon, D. Boichard, et al. Fine Mapping of Quantitative Trait Loci Affecting Female Fertility in Dairy Cattle on BTA03 Using a Dense Single-Nucleotide Polymorphism Map Genetics, April 1, 2008; 178(4): 2227 - 2235. [Abstract] [Full Text] [PDF]

				S. Mikawa, T. Hayashi, M. Nii, S. Shimanuki, T. Morozumi, and T. Awata Two quantitative trait loci on Sus scrofa chromosomes 1 and 7 affecting the number of vertebrae J Anim Sci, October 1, 2005; 83(10): 2247 - 2254. [Abstract] [Full Text] [PDF]

				J.-J. Kim, M. F. Rothschild, J. Beever, S. Rodriguez-Zas, and J. C. M. Dekkers Joint analysis of two breed cross populations in pigs to improve detection and characterization of quantitative trait loci J Anim Sci, June 1, 2005; 83(6): 1229 - 1240. [Abstract] [Full Text] [PDF]

				M. GEORGES and L. ANDERSSON Positional Identification of Structural and Regulatory Quantitative Trait Nucleotides in Domestic Animal Species Cold Spring Harb Symp Quant Biol, January 1, 2003; 68(0): 179 - 188. [Abstract] [PDF]