Genetics / statistical genomics · BEGINNER LESSON
Fine-mapping
Fine-mapping is a statistical method used after a genome-wide association study (GWAS) to narrow a genomic region to a smaller set of candidate causal variants. It uses association statistics and linkage disequilibrium (LD), the correlation pattern among nearby genetic variants, to estimate which variant or variants are most likely causal.
What fine-mapping is and what it measures
A genome-wide association study, or GWAS, measures how strongly genetic variants are associated with a trait. Fine-mapping is the next step: it asks which variant or variants inside an associated region are most likely to be the causal ones. The output is usually a probability for each variant, such as a posterior inclusion probability (PIP), meaning the estimated chance that a variant belongs to the set of causal variants. Some methods also report credible sets, which are groups of variants expected to contain the true causal variant with high probability.
[1] Fine-mapping a genome-wide meta-analysis of 98,374 migraine cases identifies 181 sets of candidate causal variants [2] Genome-wide fine-mapping improves identification of causal variants [3] FINEMAP-miss: fine-mapping genome-wide association studies with ...
How the method works
Fine-mapping compares the observed association statistics with the LD structure in the region. LD means nearby variants are correlated, so a non-causal variant can appear associated because it tracks with a causal one. Bayesian methods such as FINEMAP, SuSiE, and CAVIAR use GWAS summary statistics plus LD information to estimate how plausible each variant is as causal. Some newer approaches analyze many regions or even the whole genome and can also use functional annotations, which are external biological labels about variants.
[1] Fine-mapping a genome-wide meta-analysis of 98,374 migraine cases identifies 181 sets of candidate causal variants [2] Genome-wide fine-mapping improves identification of causal variants [3] FINEMAP-miss: fine-mapping genome-wide association studies with ...
Why the input data matter
The best fine-mapping uses in-sample LD, meaning LD calculated from the same original GWAS data. When those genotype data are unavailable, researchers often use reference LD from another panel. That approximation can create errors, especially when the reference LD does not match the study data well. Meta-analyses are especially difficult because they combine multiple studies, and no single analyst usually has the exact LD for the full combined sample. Missing variants can also matter, because excluding variants that are not present in all component studies may remove true causal variants.
[1] Fine-mapping a genome-wide meta-analysis of 98,374 migraine cases identifies 181 sets of candidate causal variants [3] FINEMAP-miss: fine-mapping genome-wide association studies with ...
What the results can and cannot tell you
Fine-mapping can narrow a large associated region to a small set of candidates, but it does not directly prove causality. Its probabilities depend on the quality of the association data, LD matching, variant coverage, and model assumptions. In migraine examples, some regions were mapped reliably, but others were less reliable because reference LD differed from in-sample LD or because important variants were missing. Newer methods try to improve calibration, meaning the reported probabilities better match the true chance that a variant is causal, but performance still depends on the data and setting.
[1] Fine-mapping a genome-wide meta-analysis of 98,374 migraine cases identifies 181 sets of candidate causal variants [2] Genome-wide fine-mapping improves identification of causal variants [3] FINEMAP-miss: fine-mapping genome-wide association studies with ...
SEE THE IDEA
How fine-mapping narrows a GWAS signal
Finds a genomic region where variants are linked to a trait.
Shows which nearby variants are correlated and may be confounded.
Combines association statistics and LD to score each variant.
Outputs a ranked shortlist, often with probabilities or credible sets.
GWAS identifies a region; fine-mapping uses LD and statistics to narrow that region to the most likely causal variants.
Fine-mapping is a narrowing step, not a direct experimental proof.
TRY IT YOURSELF
Test your understanding
Choose an answer, then check the explanation.
PUT IT TOGETHER
Match each concept to its meaning
Drag a label onto its description, or select a label and then select a description. This also works with a keyboard.
4 pairs to match
Sources and further reading
This general lesson was generated from public source material. It explains a concept; it does not interpret a private paper or provide medical advice.