BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Statistical image processing for electron microscopy on molecular 
 machines - Sjors Scheres\, MRC Laboratory of Molecular Biology
DTSTART:20110613T130000Z
DTEND:20110613T143000Z
UID:TALK31101@talks.cam.ac.uk
CONTACT:Richard Samworth
DESCRIPTION:We use three-dimensional electron microscopy to visualise mole
 cular machines\,\nwhich are protein and/or RNA complexes that fulfill vita
 l processes in any living\norganism. Despite their small size molecular ma
 chines often function in ways that\nare strikingly similar to machines fro
 m daily life\, forming the nano-scale\nequivalents of water mills\, fuel-d
 riven motors or even walking legs. In the\nelectron microscope two-dimensi
 onal projection images of individual complexes are\nrecorded. Multiple ima
 ges from complexes in different orientations are then\ncombined to obtain 
 a three-dimensional reconstruction. However\, the recorded\nimages are ext
 remely noisy (with typical signal-to-noise ratios below 0.1) and\ntheir re
 lative orientations are unknown. Consequently\, 3D-reconstruction of\nelec
 tron microscopy images is a severely ill-posed problem with incomplete dat
 a.\nI will discuss statistical image processing approaches that address th
 is problem\,\nin particular a maximum-likelihood approach that marginalise
 s over the unknown\norientations [1]\, and a maximum-a-posteriori approach
  that prevents overfitting\nthrough the use of a regularised likelihood fu
 nction (in preparation).\n\n[1] F.J. Sigworth\, P. Doerschuk\, J.M. Carazo
 \, S.H.W. Scheres (2010) "An\nintroduction to maximum-likelihood methods i
 n cryo-EM"\, Methods in Enzymology\,\n482\, 263-294.  http://dx.doi.org/10
 .1016/S0076-6879(10)82011-7\n
LOCATION:MR11\, CMS
END:VEVENT
END:VCALENDAR