Molecular Marker Evaluation
Molecular markers (or marker loci) are DNA sequence-based landmarks for identifying chromosome segments and constructing linkage maps. To be useful, a marker locus must be polymorphic in a specific mapping population, i.e., it must exist in two or more distinguishable allelic forms.
Marker types typically used in plant QTL studies are known by their acronyms: SSR (aka microsatellite), RFLP, AFLP, and RAPD, all defined in the Glossary. Whichever type of marker is used in a QTL study, the objectives are
1) To obtain complete genome coverage, i.e., to include markers located in all regions of each chromosome.
2) To develop a genome map with markers spaced about 15 to 20 cM apart. More densely spaced markers provide little additional information (Kearsey and Farquhar, 1998).
Parental screening
Markers are first used to ’screen’ (or evaluate) the parents of a mapping population for polymorphisms, detectable differences in marker patterns. Polymorphisms can take the form of size differences of DNA bands, or presence vs. absence of a DNA band. Only markers showing a clear polymorphism between the parents will be useful in mapping the population. Fig. 5 shows the results of screening two wheat parents (K=Kauz, M=MTRWA116) for microsatellite markers. The number of polymorphic markers needed for a QTL study will depend on genome size of the plant species, the average spacing between markers, and the objectives of the study. Typically, from 75 to 200 markers are used.
Evaluating the population for markers
After polymorphic markers are identified, they are used to evaluate each line or individual of the mapping population. Each line is scored for having the marker pattern corresponding to one or the other parent, as shown in Fig. 6.
Fig. 5: Screening results for six microsatellite markers in parents of a wheat mapping population. Parental DNA was amplified in a PCR reaction using marker-specific primers, then separated on a polyacrylamide gel. K=Kauz; M=MTRWA116. Blue bars indicate clearly scorable polymorphisms; the red bar indicates a subtle polymorphism that would be difficult to score; the green bar indicates a potential presence/absence polymorphism. A molecular weight standard is on the left side of the gel, with band sizes (in base pairs) indicated.
Fig. 6: DNA banding patterns for marker GWM340 in a subset of recombinant inbred lines (RILs) developed from the wheat cross Kauz/MTRWA116. RILs are scored as having either the Kauz (A) allele or the MTRWA116 (B) allele.
Do you want to try scoring markers yourself? Practice scoring an agarose gel in an interactive animation or, if the animation does not work on your device, see the gel scoring video below.
*This animation has no audio.*
Segregation distortion
Once marker data are obtained, they are generally analyzed for evidence of segregation distortion, the deviation of observed segregation ratios from the ratios expected with Mendelian inheritance. The expected ratios will differ depending on the type of population and type of marker used. With a RIL population, all markers are expected to segregate 1 parent A : 1 parent B, i.e., both homozygous parental genotypes are expected to be present in equal proportions*. However, if the marker scores turn out to be 70% A and only 30% B, this indicates distorted segregation; for some reason, the parent A allele is transmitted to the progeny at higher frequencies than the parent B allele.
To determine whether the observed data differ significantly from expected ratios, a chi-square statistical test is performed An example of chi-square calculations to test segregation distortion is presented on the Oregon Wolfe Barley web site http://barleyworld.org/sites/default/files/2_mendel_2.pdf, under Example 1, Monohybrid Model.
*Technically, a few heterozygous individuals are also expected, depending on the generation of inbreeding from which the RILs were derived. The expected percent of heterozygosity in the F2 generation is 50%, and this decreases by half for each generation of inbreeding. Thus, for F6-derived RILs, the expected proportion of heterozyotes is 3.125%. For F8-derived RILs, the proportion decreases to 0.78%. In an F2 population, when using a codominant marker like SSR or RFLP, the expected ratio is 1 parent A : 2 heterozygotes : 1 parent B.
Linkage mapping
Scores from all markers are organized in a data file in a format that can be imported into a linkage mapping program (Fig. 7). The most commonly used software for linkage mapping is Mapmaker (Lander et al., 1987, link updated 10-25-2013), available at http://www.broadinstitute.org/ftp/distribution/software/mapmaker3/
The principles of linkage and genetic map construction are described in two separate lessons of the Library of Crop Technology, Linkage Parts 1 and 2. Viewers may consult those lessons for a detailed explanation of these topics. Suffice it to say here that the objective of linkage mapping is to develop an ordered, linear representation of marker loci for each chromosome, based upon the observed recombination rates between loci.
There are two outputs from the marker evaluation component of a QTL study:
- A file of marker scores, as shown in Fig. 7.
- A genetic linkage map, an example of which is shown in Fig. 8.
Fig. 7: Partial view of a MAPMAKER input file. The top row indicates the type of population. The second row states number of progeny (92), number of markers (146), and number of traits (0). The following rows are scores for the named markers, all preceded by an asterisk. A and B are parental scores, and a hyphen indicates missing data.
Fig. 8: Partial linkage map of wheat chromosome 2A, after Gupta et al. (2002). Marker names are listed to the right of the chromosome and linkage distances between adjacent markers, in centiMorgans, are listed to the left.