I think it is important to stress that all these different types of heritability studies are consistent. They are measuring slightly different things. The twin and family-based studies are estimating the total heritability—that is, heritability from any cause. The SNP-based studies are only able to measure the component of that heritability accounted for by common SNP-based variation. Because the DNA microarrays used for GWAS only capture common genetic variation, they cannot provide information on rare variants or other genetic variability. In other words, we have a simple sum: Total heritability = heritability due to common SNP variation + heritability due to other genetic variation. The family and twin studies are estimating the total heritability, while the SNP-based studies estimate the component due to common SNPs. The remaining unexplained heritability is likely to be due to rare genetic variation or common variants not well captured by arrays such as copy-number changes or microsatellite variability.
The problem is in identifying the relevant rare variants with confidence. If they really are rare, we will not easily be able to use standard statistical tests of association, and if they are in people with no known family history, we will not be able to test segregation within families. This is the next great challenge for ALS genetics.
Comments
Kings College London
I think it is important to stress that all these different types of heritability studies are consistent. They are measuring slightly different things. The twin and family-based studies are estimating the total heritability—that is, heritability from any cause. The SNP-based studies are only able to measure the component of that heritability accounted for by common SNP-based variation. Because the DNA microarrays used for GWAS only capture common genetic variation, they cannot provide information on rare variants or other genetic variability. In other words, we have a simple sum: Total heritability = heritability due to common SNP variation + heritability due to other genetic variation. The family and twin studies are estimating the total heritability, while the SNP-based studies estimate the component due to common SNPs. The remaining unexplained heritability is likely to be due to rare genetic variation or common variants not well captured by arrays such as copy-number changes or microsatellite variability.
The problem is in identifying the relevant rare variants with confidence. If they really are rare, we will not easily be able to use standard statistical tests of association, and if they are in people with no known family history, we will not be able to test segregation within families. This is the next great challenge for ALS genetics.
Make a Comment
To make a comment you must login or register.