196
J. D. Wall
et al. 2004). However, an analysis of the initial draft Neanderthal genome showed
that Neanderthals were more closely related to all non-African populations than to
sub-Saharan African populations (Green et al. 2010). The most likely explanation
for this is that Neanderthals and the ancestors of non-African populations interbred
and exchanged genes, perhaps in the Middle East 60–90 Kya (Sankararaman et
al. 2012), leading to greater genetic similarity. This result was initially surprising,
since researchers initially assumed that any interbreeding between Neanderthals and
modern humans would have occurred in Europe 30–40 Kya. Instead, more detailed
analyses show that East Asian individuals tend to have greater Neanderthal ancestry
than do European ones (e.g., Wall et al. 2013). While this could be due to weaker
purifying selection in East Asian populations (Sankararaman et al. 2014), it more
likely reflects a separate Neanderthal admixture event after the initial divergence of
European and East Asian populations or a recent dilution of Neanderthal ancestry
in the ancestors of European populations (Kim and Lohmueller 2015; Vernot and
Akey 2015). Thus far, a separate Neanderthal admixture event seems more likely
(see, e.g., Villanea and Schraiber 2019).
Similar studies using the draft Denisovan genome also turned up a surprising
result—aboriginal Australians, Melanesians, and Philippine Negrito groups derived
4–5% of their genome from Denisovans, whereas all other human populations
tested showed much smaller levels of Denisovan ancestry (e.g., Reich et al. 2010,
2011; Sankararaman et al. 2016; Vernot et al. 2016; GenomeAsia 100K Consortium
2019). This strongly suggests that the colonization of Melanesia involved a separate
migration out of Africa than did the colonization of mainland Eurasia, perhaps
along the Southern coast of Asia (the “Southern route” hypothesis, reviewed
by Oppenheimer 2009). If so, the historical range of the Denisovans must have
stretched far south of the Siberian cave where the fossil remains were found. There
is also growing evidence for multiple separate Denisovan admixture events in East
and Southeast Asia (Browning et al. 2018; GenomeAsia 100K Consortium 2019;
Jacobs et al. 2019).
Indirect Evidence for Ancient Admixture While the direct comparison between
modern human DNA and archaic human DNA described above has been extremely
informative, it has been limited by the scarcity of archaic human fossils with
sufficient amounts of ancient DNA. Other, indirect methods are needed to detect
potential ancient admixture between modern humans and other archaic human
groups in East Asia and Africa. These methods rely on the observation that ancient
admixture, if it occurred, will leave large, discrete chunks of introgressed sequence
that are substantially different from orthologous modern human sequences. These
chunks can then be detected by searching for unusual patterns of LD (e.g., Wall
2000; Plagnol and Wall 2006; Wall et al. 2009), in an analogous way to how local
ancestry is estimated across the genomes of recently admixed individuals (e.g., Price
et al. 2009; Baran et al. 2012). Analyses of human polymorphism data have found
evidence for ancient admixture in sub-Saharan African populations (Garrigan et al.
2005; Hayakawa et al. 2006; Wall et al. 2009; Hammer et al. 2011; Lachance et
al. 2012; Hsieh et al. 2016; Durvasula and Sankararaman 2020; Wall et al. 2019),
J. D. Wall
et al. 2004). However, an analysis of the initial draft Neanderthal genome showed
that Neanderthals were more closely related to all non-African populations than to
sub-Saharan African populations (Green et al. 2010). The most likely explanation
for this is that Neanderthals and the ancestors of non-African populations interbred
and exchanged genes, perhaps in the Middle East 60–90 Kya (Sankararaman et
al. 2012), leading to greater genetic similarity. This result was initially surprising,
since researchers initially assumed that any interbreeding between Neanderthals and
modern humans would have occurred in Europe 30–40 Kya. Instead, more detailed
analyses show that East Asian individuals tend to have greater Neanderthal ancestry
than do European ones (e.g., Wall et al. 2013). While this could be due to weaker
purifying selection in East Asian populations (Sankararaman et al. 2014), it more
likely reflects a separate Neanderthal admixture event after the initial divergence of
European and East Asian populations or a recent dilution of Neanderthal ancestry
in the ancestors of European populations (Kim and Lohmueller 2015; Vernot and
Akey 2015). Thus far, a separate Neanderthal admixture event seems more likely
(see, e.g., Villanea and Schraiber 2019).
Similar studies using the draft Denisovan genome also turned up a surprising
result—aboriginal Australians, Melanesians, and Philippine Negrito groups derived
4–5% of their genome from Denisovans, whereas all other human populations
tested showed much smaller levels of Denisovan ancestry (e.g., Reich et al. 2010,
2011; Sankararaman et al. 2016; Vernot et al. 2016; GenomeAsia 100K Consortium
2019). This strongly suggests that the colonization of Melanesia involved a separate
migration out of Africa than did the colonization of mainland Eurasia, perhaps
along the Southern coast of Asia (the “Southern route” hypothesis, reviewed
by Oppenheimer 2009). If so, the historical range of the Denisovans must have
stretched far south of the Siberian cave where the fossil remains were found. There
is also growing evidence for multiple separate Denisovan admixture events in East
and Southeast Asia (Browning et al. 2018; GenomeAsia 100K Consortium 2019;
Jacobs et al. 2019).
Indirect Evidence for Ancient Admixture While the direct comparison between
modern human DNA and archaic human DNA described above has been extremely
informative, it has been limited by the scarcity of archaic human fossils with
sufficient amounts of ancient DNA. Other, indirect methods are needed to detect
potential ancient admixture between modern humans and other archaic human
groups in East Asia and Africa. These methods rely on the observation that ancient
admixture, if it occurred, will leave large, discrete chunks of introgressed sequence
that are substantially different from orthologous modern human sequences. These
chunks can then be detected by searching for unusual patterns of LD (e.g., Wall
2000; Plagnol and Wall 2006; Wall et al. 2009), in an analogous way to how local
ancestry is estimated across the genomes of recently admixed individuals (e.g., Price
et al. 2009; Baran et al. 2012). Analyses of human polymorphism data have found
evidence for ancient admixture in sub-Saharan African populations (Garrigan et al.
2005; Hayakawa et al. 2006; Wall et al. 2009; Hammer et al. 2011; Lachance et
al. 2012; Hsieh et al. 2016; Durvasula and Sankararaman 2020; Wall et al. 2019),