Pooling the samples

[lindenb]

We're looking for a variation causing a rare disease.

This variation should be a monogenic, dominant, autosomal mutation.

Would it be worth pooling the DNA of some affected individuals of the same family before running an exome sequencing ?

The non-causal mutations would be diluted isn't it ?
What would be the problems ?

Many thanks

[adamdeluca]

Probably a bad idea to pool in this case.

Without pooling you are looking for a 50% allele (het), with pooling you would still be looking for a 50% allele (doesn't help to detect). In addition, with the pooling you will have 50% alleles that are 1/2 of the affected are homozygous for allele A and half for B (not biologically interesting).

Non-causal mutations would not be diluted very much because these are family members and you would expect them to share large pieces of the genome.

I think the best way to use families is to run SNP chips and sequence one person. The chips even in families too small to get significant linkage can help to eliminate regions.

[lindenb]

Hi Adam,
thank you very much for your answer.

Tell me if I'm wrong, but if I pool the DNA of a child, his father, and his grand-father (all are affected), I will get 50% alleles with the causal mutation.

For another loci:

- the 3 individuals are homozygous: I discard this mutation because the causal allele is rare.

- the 3 individuals have the very same combination of alleles children=A1/A2 father=A1/A2 g:father=A1/2 , then, as you said, the signal would be 50% just like with a non-pooled assay

- if there is more than two alleles then the calling would be a mixture of bases with a lower calling quality This kind of non informative mutation would be diluted. What kind of extra information would I miss if I was not pooling those DNAs ?

or am I missing something... ?

[adamdeluca]

c: A1/A1
f: A1/A2
gf: A2/A2

would also be 50%.