Runs of homozygosity

Question

[Sharon]

What does overlapping ROH with allelic match mean as defined in plink? I mean what we can tell from overlapping ROH with/without allelic match? Does this refer to anything to kinship relatedness? Like if we have two persons who have long overlapping ROH with allelic match, does this mean they are possibly related?

Thanks

Answer 1

[Kevin Blighe]

Allelic matching is an important parameter to consider when determining relatedness via runs of homozygosity (ROH). The general threshold (for relateness or identity by descent) in any defined genomic region is that > 95% of all homozygous calls in the region should have the same call in the 2 or more individuals being studied (i.e., the homozygous calls should match allelically). If only 10% match, then it's not strong evidence that the region of DNA in question was inherited equally amongst the individuals being studied, even if it's a ROH, and implies genetic inheritance outside of those being studied.

However, as you can imagine, the metric is heavily dependent on the size of the region in question, SNP genotyping density, the number of heterozygous calls permitted (and assumed to be sequencing errors), and the number of missing calls.

If you have 2 individuals with long (>1Mbp?) ROH that allelically match, then it's evidence (but not proof) that they exhibit relatedness. I studied a family once in which one offspring was homozygous for a very rare variant (0.0002% frequency in 1000 Genomes). A check of ROH in the family revealed that the parents were possibly closely related by descent.

[Sharon]

Thank you. It would help me more if you can give me an example of ROH with allelic match versus without allelic match?

[Kevin Blighe]

Sure, if you look at this example, assuming that these are samples 'randomly' taken from a small population of people living in a remote and mostly inaccessible location, all samples exhibit ROH across the 6 SNPs studied. However, only Sample1, Sample2, and Sample4 allelically match.

         SNP1   SNP2   SNP3   SNP4   SNP5   SNP6
Sample1  A/A    T/T    T/T    C/C    G/G    G/G
Sample2  A/A    T/T    T/T    C/C    G/G    G/G
Sample3  C/C    T/T    A/A    C/C    A/A    G/G
Sample4  A/A    T/T    T/T    C/C    G/G    G/G

[Sharon]

Got it. Thanks so much. Many thanks