9 Natural Selection, Genetic Variation, and Human Diversity
207
be identified by molecular evidence (Tishkoff et al. 2001; Sabeti et al. 2002; Verrelli
et al. 2002).
Frequency-dependent selection is common in host-pathogen interactions, and it
is known to be an important force acting on alleles at the human leukocyte antigen
(HLA) genes, which will be discussed in detail below. A recent study examined
SNP data in European and African American samples and found evidence for
long-term, frequency-dependent selection at 60 genes, including immunity genes,
keratins, membrane channels, and cellular structure genes (Hahn 2008; Andrés et al.
2009). Although balancing selection undoubtedly plays an important role in shaping
variation at perhaps dozens of genes, it is likely the rarest type of selection acting
on the human genome (Akey 2009; Andrés et al. 2009).
9.2.2 Negative Selection
The effects of negative selection are difficult to quantify for a single locus, so
studies have focused on describing its impact genome-wide (Reed et al. 2005;
Charlesworth 2006; McVicker et al. 2009; Lohmueller et al. 2011a). Unexpectedly
low levels of diversity near functional elements would provide evidence for ongoing
negative selection in the human genome, and this pattern is borne out by studies
showing reduced variation near evolutionarily conserved elements (Schroeder et al.
1995; McVicker et al. 2009; Lohmueller et al. 2011a). Variation is even further
reduced near human exons, and estimates show that selection has reduced variation
by 19–26% on the autosomes and by 12–40% on the X chromosome (Krawczak
and Zschocke 2003; McVicker et al. 2009). Although hitchhiking caused by
positive selection also makes a contribution to this reduced variation, widespread
weak purifying selection combined with positive selection in a small fraction of
the genome can explain the observed patterns (Allison 1954; Reed et al. 2005;
McVicker et al. 2009; Quintana-Murci and Barreiro 2010; Lohmueller et al. 2011a).
We can assume, therefore, that negative selection is acting on a large proportion of
the functional elements in the human genome.
9.2.3 Positive Selection
Much debate has centered on the relative contribution of purifying and positive
selection affecting genetic variation (Verrelli et al. 2002; Stephan 2010). Current
evidence suggests that weak negative selection is widespread, whereas positive
selection has affected a smaller portion of the genome more strongly. As a test
of this hypothesis, Hernandez et al. searched for a reduction of neutral genetic
diversity surrounding amino acid human substitutions. This is another signature
of selective sweeps (Fig. 9.1a). They found that the reduction in neutral diversity
surrounding amino acid substitutions was similar to that surrounding putatively
neutral substitutions. Thus, they concluded that fewer than 10% of amino acid
substitutions in the human lineage were fixed by selective sweeps. A more recent
207
be identified by molecular evidence (Tishkoff et al. 2001; Sabeti et al. 2002; Verrelli
et al. 2002).
Frequency-dependent selection is common in host-pathogen interactions, and it
is known to be an important force acting on alleles at the human leukocyte antigen
(HLA) genes, which will be discussed in detail below. A recent study examined
SNP data in European and African American samples and found evidence for
long-term, frequency-dependent selection at 60 genes, including immunity genes,
keratins, membrane channels, and cellular structure genes (Hahn 2008; Andrés et al.
2009). Although balancing selection undoubtedly plays an important role in shaping
variation at perhaps dozens of genes, it is likely the rarest type of selection acting
on the human genome (Akey 2009; Andrés et al. 2009).
9.2.2 Negative Selection
The effects of negative selection are difficult to quantify for a single locus, so
studies have focused on describing its impact genome-wide (Reed et al. 2005;
Charlesworth 2006; McVicker et al. 2009; Lohmueller et al. 2011a). Unexpectedly
low levels of diversity near functional elements would provide evidence for ongoing
negative selection in the human genome, and this pattern is borne out by studies
showing reduced variation near evolutionarily conserved elements (Schroeder et al.
1995; McVicker et al. 2009; Lohmueller et al. 2011a). Variation is even further
reduced near human exons, and estimates show that selection has reduced variation
by 19–26% on the autosomes and by 12–40% on the X chromosome (Krawczak
and Zschocke 2003; McVicker et al. 2009). Although hitchhiking caused by
positive selection also makes a contribution to this reduced variation, widespread
weak purifying selection combined with positive selection in a small fraction of
the genome can explain the observed patterns (Allison 1954; Reed et al. 2005;
McVicker et al. 2009; Quintana-Murci and Barreiro 2010; Lohmueller et al. 2011a).
We can assume, therefore, that negative selection is acting on a large proportion of
the functional elements in the human genome.
9.2.3 Positive Selection
Much debate has centered on the relative contribution of purifying and positive
selection affecting genetic variation (Verrelli et al. 2002; Stephan 2010). Current
evidence suggests that weak negative selection is widespread, whereas positive
selection has affected a smaller portion of the genome more strongly. As a test
of this hypothesis, Hernandez et al. searched for a reduction of neutral genetic
diversity surrounding amino acid human substitutions. This is another signature
of selective sweeps (Fig. 9.1a). They found that the reduction in neutral diversity
surrounding amino acid substitutions was similar to that surrounding putatively
neutral substitutions. Thus, they concluded that fewer than 10% of amino acid
substitutions in the human lineage were fixed by selective sweeps. A more recent