Advancedpopulation
Linkage Analysis
Overview
Linkage analysis maps disease genes by tracking marker-disease co-segregation in families: markers near the disease gene travel together (few recombinants). The LOD score (log10 of odds favoring linkage) quantifies evidence: LOD >= 3 (1000:1) declares linkage. This mapped Huntington's (1983, chromosome 4), CFTR, and BRCA1 before the genome era; GWAS and sequencing have largely superseded it for complex traits.
Learning objectives
- Explain co-segregation logic
- Interpret LOD scores
- Describe how Huntington's was mapped
- Compare linkage vs association
The logic
In a family, a DNA marker physically near the disease gene recombines rarely with it: affected relatives share the same marker allele (identical by descent). Count recombinants vs non-recombinants across meioses: few recombinants -> linked. Recombination fraction theta estimates distance.
Markers and genes traveling together through pedigrees must be neighbors.
LOD scores
LOD = log10(likelihood of data if linked at theta / if unlinked). LOD 3 = 1000:1 odds -> significant (Morton, 1955). LOD -2 excludes linkage. Multipoint analysis uses many markers simultaneously; affected-sib-pair methods work without knowing the mode of inheritance.
The traditional threshold balancing false positives across the genome.
Triumphs
Huntington's (1983): first disease gene mapped by linkage (Gusella, chromosome 4p), a decade before the gene was found. CFTR (1989), BRCA1 (1994), and hundreds of Mendelian genes fell to linkage + positional cloning. It owned the 1985-2005 gene-mapping era.
Linkage excelled at single-gene disorders but lacked power for small-effect polygenes.
Linkage vs association
Linkage: families, rare variants, large effects, coarse resolution (megabases). Association (GWAS): populations, common variants, small effects, fine resolution. Sequencing now finds Mendelian genes directly; linkage remains useful for mapping in large pedigrees and validating.
Linkage found the big-effect genes; GWAS found the small-effect multitudes.
Key points
- Markers co-segregating with disease across meioses indicate linkage
- LOD >= 3 declares linkage (1000:1 odds)
- Mapped Huntington's (1983), CFTR, BRCA1 pre-genome
- Linkage: families/rare/large-effect; GWAS: populations/common/small-effect
Common misconceptions
✓ Truth: LOD is log10 odds: 3 = 1000:1 odds favoring linkage, roughly p ~ 10^-4.
✓ Truth: It maps to megabase regions; positional cloning/sequencing then finds the gene.
Practice questions
Q1. A LOD score of 3 means
- p = 0.03
- 1000:1 odds in favor of linkage
- 3% recombination
- 3 genes
log10(1000) = 3: the traditional significance threshold.
Q2. Linkage analysis tracks ___ in families
- Mutations directly
- Marker-disease co-segregation
- Proteins
- Fossils
Shared marker alleles among affecteds imply a nearby disease gene.
Q3. Huntington's disease was mapped to chromosome 4 in
- 1953
- 1983
- 2003
- 2023
Gusella's landmark: the first disease gene mapped by linkage.
Click a question to reveal the answer and explanation.
Frequently asked questions
Why did Huntington's mapping matter?
First autosomal disease gene mapped (1983), enabling predictive testing a decade before gene identification.
What is multipoint linkage?
Using multiple markers simultaneously to extract maximal inheritance information across a region.
Why did linkage fail for diabetes?
Complex traits involve many small-effect variants: linkage lacks power where GWAS (barely) succeeds.