Live Hot, Die Young: Transmission distortion in Recombination Hotspots

1
Live Hot, Die YoungTransmission distortion in
Recombination Hotspots

• Graham Coop and Simon Myers
• University of Chicago
• and University of Oxford

2

• Hotspots are not conserved between human and
  chimp
• Wall et. al. (03), Ptak et. al. (04), Ptak et.
  al. (05), Winckler and Myers et. al. (05)
• One explanation of this is that alleles that
  locally disrupt hotspots are subject to over
  transmission
• Investigated population genetics models and
  consequences of this fact

3
What drives us?

• Probability of survival of an ancestral hotspot
• Investigate the spectrum of hotspot heats under
  different models of how hotspots arise
• Talk about the patterns left in linkage
  disequilibrium by segregating hotspots

4
Crossing over and Gene conversion are thought to
be the result of Double Stranded Breaks (DSBs)
5
DSB occurs
6
Genetic information in region surrounding the
break is lost
Gene Conversion is used to 'copy' information
from other chromosome
7
Crossing over and gene conversion
gene conversion only
8
If in a heterozygote
DSBs are initiated more on the A background than
the B background
Then the A allele is under transmitted and The B
allele is over transmitted
This has been observed at the DNA2 hotspot in
Humans (Jeffreys and Neumann 02)
9
Driven to Destruction

• Alleles that locally disrupt a hotspot will be
  driven through the population by this biased
  transmission
• This leads to the recombination hotspot paradox
• How do hotspots persist in the face of this
  drive?
• Bullard et. al. (97), Pineda-Krch and Redfield
  (05)

10
Biased gene conversion

• Let
• rH the difference in the DSBs initiation rate
  between the hotspot and nonhotspot allele
• p the probability that the allele that
  initiated the DSB is transmitted (p
• Biased gene conversion acts exactly like
  additive selection (Nagylaki, 83), thus if the
  current frequency of the disrupting allele is x
  then
• E(?x) 2rH(1/2-p)x(1-x)
• scaling by 2Ne gives us the population scaled
  drive

11
MHC sperm crossing over map
DNA3
DNA2
DNA1
12
Survival between human and chimps

• A hotspot is most likely to survive between
  species if it is fixed in the ancestral species
• A hotspot will survive to the current day if NO
  disrupting mutations have arisen and fixed in
  either population
• Affected by two factors
• Number of sites where disrupting mutations can
  arise
• Heat of a hotspot - determines the probability
  that a disrupting mutation fixes

13
Survival between human and chimps
14
Hotspot Genesis
15
Hotspot Genesis
Assuming that the hotspots of different heats
arise at the same rate, we can find expectation
of number of hotspots currently above a
particular frequency (in this case 0.5)
16
Hotspot Genesis
New hotspot alleles may not experience drive
against themselves, because they promote -
DSBs at some distal position, or - DSBs on
both chromosomes equally Hotspots may still be
removed from the population by locally acting
disrupting mutations We find that the number of
hotspots of heat, rH, behaves approximately as
1/(NerH)
17
Live Fast, Die Young, and leave no great looking
corpse
You might expect that an allele that locally
turns on/off a hotspot would result in a
background specific pattern of linkage
disequilibrium By analogy to the genealogy of a
selected mutation (Hudson and Kaplan 88) we
constructed a coalescent process for a
segregating hotspot promoting/disrupting allele
We (annoyingly) find that such alleles do not
result in background specific patterns of LD, and
so will in general be undetectable
18
Conclusions

• Hot hotspots are unlikely to survive even if
  there are few sites that can affect the heat
• Hotspots are likely to arise in a way that
  initially shields them from drive
• Alleles that affect the heat of hotspots will
  leave little (if any) signal in LD data
• We've also looked at how selection for
  recombination due to correct segregation at
  meiosis affects the number of hotspots on a
  chromosome