Statistical genetics · BEGINNER LESSON
In-sample LD
In-sample LD is linkage disequilibrium measured directly in the same study sample, or in a random subset of that sample, that produced the GWAS summary statistics. It is used to match the correlation structure of the data more closely than an outside reference panel.
What it is
Linkage disequilibrium, or LD, is non-random correlation between genetic variants. In LD score methods, what is measured is how strongly each single-nucleotide polymorphism, or SNP, correlates with nearby SNPs. In-sample LD means those correlations are estimated from the same people whose genome-wide association study, or GWAS, summary statistics are being analyzed, or from a random subset of them. This can be done with covariate-adjusted genotypes when the summary statistics were also adjusted for the same covariates such as principal components.
[1] Estimating heritability and its enrichment in tissue-specific gene sets ... [2] Accurate and efficient estimation of local heritability using summary statistics and the linkage disequilibrium matrix - Nature Communications [3] LDER-GE estimates phenotypic variance component of gene–environment interactions in human complex traits accurately with GE interaction summary statistics and full LD information
How it works
The basic idea is to use the study sample itself so the LD matches the data generating the summary statistics. For LD score regression, the LD score of a SNP is the sum of pairwise correlations squared with other SNPs. In populations with mixed ancestry, using LD from an unmatched reference panel can distort estimates because long-range correlations from admixture may be present. The sources report that in-sample LD, especially when adjusted for the same covariates as the GWAS, gives more accurate and less biased results, and that a random subset can be enough in practice. One study recommends at least about 1000 individuals for computing in-sample LD scores in that approach.
[1] Estimating heritability and its enrichment in tissue-specific gene sets ... [2] Accurate and efficient estimation of local heritability using summary statistics and the linkage disequilibrium matrix - Nature Communications
Why it matters
Using in-sample LD supports summary-statistic methods that estimate heritability, enrichment in tissues or cell types, and related variance components without needing every individual’s raw data. The sources also note that accurate in-sample LD can be computed from a subset, which helps large cohorts. Another method using full LD information reports improved precision over methods that use only simplified LD summaries, showing why the choice of LD input affects statistical efficiency.
[1] Estimating heritability and its enrichment in tissue-specific gene sets ... [2] Accurate and efficient estimation of local heritability using summary statistics and the linkage disequilibrium matrix - Nature Communications [3] LDER-GE estimates phenotypic variance component of gene–environment interactions in human complex traits accurately with GE interaction summary statistics and full LD information
Limitations
In-sample LD is not a universal fix. It still depends on how well the study sample represents the data being analyzed, and methods can be sensitive if the LD is approximated too aggressively or if the sample is small. One source warns that low minor allele frequency, or MAF, SNPs can reduce performance when the reference panel is small. Another source notes that summary-statistic methods may still rely on assumptions such as polygenic effects, and that some approaches depend on a full in-sample LD matrix.
[1] Estimating heritability and its enrichment in tissue-specific gene sets ... [2] Accurate and efficient estimation of local heritability using summary statistics and the linkage disequilibrium matrix - Nature Communications [3] LDER-GE estimates phenotypic variance component of gene–environment interactions in human complex traits accurately with GE interaction summary statistics and full LD information
SEE THE IDEA
How in-sample LD is used in summary-statistic genetics
People whose genotypes and summary statistics come from the same study
Correlations among SNPs are measured inside that sample or a subset
LD scores or full LD matrices are fed into summary-statistic methods
Heritability or enrichment estimates are more closely matched to the study data
Same-sample genotype information is used to compute LD, and that LD then supports downstream summary-statistic estimation.
In-sample LD links the study data to the LD input so the correlation structure better matches the analysis.
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.
- Estimating heritability and its enrichment in tissue-specific gene sets ...
- Accurate and efficient estimation of local heritability using summary statistics and the linkage disequilibrium matrix - Nature Communications
- LDER-GE estimates phenotypic variance component of gene–environment interactions in human complex traits accurately with GE interaction summary statistics and full LD information