Linkage mapping/recombination mapping/positional cloning
- rely on known markers (typically SNPs) that are close to the gene
responsible for a disease or trait to segregate with that marker within a
family. Works great for high-penetrance, single gene traits and
QTL mapping/interval mapping - for quantitative
traits like height that are polygenic. Same as linkage mapping except
the phenotype is continuous and the markers are put into a scoring
scheme to measure their contribution - i.e. "marker effects" or "allelic
contribution". Big in agriculture.
GWAS/linkage disequilibrium mapping - score
thousands of SNPs at once from a population of unrelated individuals.
Measure association with a disease or trait with the presumption that
some markers are in LD with, or actually are, causative SNPs.
So linkage mapping and QTL mapping are similar in that they rely on
Mendelian inheritance to isolate loci. QTL mapping and GWAS are similar
in that they typically measure association in terms of log-odds along a
genetic or physical map and do not assume one gene or locus is
responsible. And finally, linkage mapping and GWAS are both concerned
with categorical traits and diseases.