BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:Europe/Vienna
BEGIN:DAYLIGHT
DTSTART:20170326T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20171029T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260427T100747Z
UID:59d21f0e5d9e5993380144@ist.ac.at
DTSTART:20171006T100000
DTEND:20171006T110000
DESCRIPTION:Speaker: Stuart J.E. Baird\nhosted by Nick Barton\nAbstract: Po
 pulation genomics requires us to summarise large quantities of information
 . Ideally this would be through lossless compression\, such that the entir
 e original information could be reconstructed from the summary. Inference 
 could then proceed by co-estimation of parameters and error rates\, avoidi
 ng the hazards of stepwise estimation. In contrast: a) Site-by-site summar
 ies lose flanking sequence context\; inefficient because a site and its fl
 anks may be embedded within a tract of shared history. b) Within-individua
 l summaries lose population context\; inefficient because histories can be
  shared across tracts within individuals. Such tracts of shared history ca
 n be compressed without loss of information using run length encoding (RLE
 ) eg: all individuals have the same history for a run of length 1234 sites
 . This suggests RLE as a potentially efficient compression alternative to 
 site-by-site/within-individual summarisations that in addition retains bot
 h sequence and  population context. Retaining this context allows better i
 nformed thresholding decisions\, such as 'calling' sequence state. The rob
 ustness of inference to arbitrarily chosen thesholding levels is much more
  efficiently explored when such decisions occur at the end\, rather than t
 he start\, of information processing. As a worked example\, I explore the 
 construction and properties of RLE-based summaries for population genomics
  using a sample of 19 mice sampled across Eurasia and the European house m
 ouse hybrid zone.
LOCATION:Mondi Seminar Room 2\, Central Building\, ISTA
ORGANIZER:abonvent@ist.ac.at
SUMMARY:Stuart J.E. Baird: Compression population genomics
URL:https://talks-calendar.ista.ac.at/events/860
END:VEVENT
END:VCALENDAR