Page 95 - Genetics_From_Genes_to_Genomes_6th_FULL_Part3

11.5 The Era of Whole-Genome Sequencing 389

by two different mutations. Finally, if an inheritance pat- Figure 11.26b presents a graphical summary of part of
tern shows sex linkage, the search for candidate genes this gigantic data set, showing the landscape of recombina-
would be limited to the X chromosome; if autosomal, the tion that occurred on just two chromosomes during meiosis
X chromosome would be excluded. in the mother and father to produce the gametes resulting in
DNA sequence information from the patient’s relatives the affected children. The hypothesis that Miller syndrome
is particularly useful in winnowing down the list of candi- is recessive predicts that the responsible gene would lie in
date polymorphisms. As an example, SNP genotyping of a region where the affected son and daughter share the
relatives using microarrays, as discussed earlier in this chap- same allele from the mother and also the same allele from
ter, could narrow the search to a region between two known the father (regions labeled identical in the figure). Geneti-
SNPs. Positional cloning and whole-genome sequencing are cists studying the disease could thus focus their attention
thus not mutually exclusive approaches to disease gene iden- only on the approximately 25% of the genome where this
tification; instead, they can provide complementary infor- was the case (Fig. 11.26b). We describe the outcome of this
mation. Better yet, though more expensive, would be investigation later on in this chapter.
comparisons of the patient’s whole-genome or exome se-
quence with those of parents and/or siblings.
A recent case study illustrates the power of DNA se- Clues from a variant’s predicted effect
quence information from related individuals (Fig. 11.26). on gene function
A brother and sister had Miller syndrome, a rare condition Researchers first try to look for disease-causing mutations
affecting development of the face and limbs, but neither in the protein coding regions of the exome because these
parent was affected (Fig. 11.26a). These facts suggest (but parts of the genome are the easiest to look at: Coding re-
don’t prove) that Miller syndrome is a recessive autosomal gions constitute only a small fraction of the total genome,
condition, with the two children inheriting mutant alleles and alterations in the coding region are the most straight-
from both of their heterozygous, carrier parents. To find forward to interpret. In particular, investigators would
the Miller syndrome gene, researchers sequenced the entire search for rare polymorphisms that change the identity of
genomes of both parents and both children; in fact, this was an amino acid (that is, SNPs causing missense mutations)
the first time in history that the genomes of all the members or alter the reading frame (SNPs, DIPs, or SSRs causing
of a nuclear family were sequenced completely. nonsense or frameshift mutations). Most nonsense/frame-
shift mutations and a subset of missense mutations will be
nonanonymous DNA polymorphisms that affect phenotype
Figure 11.26 The first family with completely sequenced through changes in protein function. In contrast, silent mu-
genomes. (a) Pedigree for Miller syndrome. (b) Map showing allele tations that change a codon into a different codon for the
inheritance along chromosomes 16 and 17 in the affected children. same amino acid will not affect phenotype; these anony-
In identical regions, the affected brother and sister share the same mous SNPs can therefore be discarded as candidates.
maternally and paternally derived alleles. In nonidentical regions, the The assumption that a particular genetic disease results
siblings share no alleles. In haploidentical maternal regions, the
siblings have the same allele from the mother but different alleles from a mutation in a protein-coding exon is nonetheless
from the father. In haploidentical paternal regions, the brother and very uncertain. Some genetic diseases are caused not by
sister share a common allele from the father but have different alterations in amino acid sequence, but rather by the amount
alleles from the mother. If Miller syndrome is recessive, the of a protein that the organism produces. Mutations that re-
responsible gene should lie in an identical region. This prediction side in regions of the genome outside of the exome could,
was upheld when mutations in the DHOD gene on chromosome 16
were found to cause the disease. for example, lower or prevent the transcription of a gene or
(a) (b) the splicing of its primary transcript. Either case would
lower the amount of the gene’s protein product or even
prevent its synthesis. Such mutations would never be found
if researchers focused their attention only on the exome.
And unfortunately, we still understand so little about the
DNA sequences that regulate transcription or splicing that
many such mutations will be overlooked even when the
patient's whole-genome sequence is available.

Clues from previously determined
genome sequences
Rare diseases are unlikely to be caused by variants com-
mon in the human population. As a result, variants that
have been documented in databases as common are poor