BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Using Gaussian process models to infer pseudotime and identify gen
 e-specific branching dynamics from single-cell data - Alexis Boukouvalas\,
  Head of Machine Learning at Prowler.io
DTSTART:20190604T130000Z
DTEND:20190604T140000Z
UID:TALK126025@talks.cam.ac.uk
CONTACT:Simon Staines
DESCRIPTION:We demonstrate how to develop and apply Gaussian Process model
 s for dimensionality reduction and inference of branching dynamics in sing
 le cell transcriptomic data. We will discuss two models: \n\nGrandPrix: an
  efficient implementation of the Gaussian process latent variable model wh
 ich allows scaling up the GP approach to modern single-cell datasets. We a
 pply our method on microarray\, nCounter\, RNA-seq\, qPCR and droplet-base
 d datasets from different organisms. The model converges an order of magni
 tude faster compared to existing methods whilst achieving similar levels o
 f estimation accuracy.\n\nThe branching Gaussian process (BGP): a non-para
 metric model that is able to identify branching dynamics for individual ge
 nes and provide an estimate of branching times for each gene with an assoc
 iated credible region. We demonstrate the effectiveness of our method on s
 imulated data\, a single-cell RNA-seq haematopoiesis study and mouse embry
 onic stem cells generated using droplet barcoding. The method is robust to
  high levels of technical variation and dropout\, which are common in sing
 le-cell data.\n
LOCATION:Large Seminar Room\, 1st floor\, Institute of Public Health\, For
 vie Site\, Cambridge BioMedical Campus\, CB2 0SR
END:VEVENT
END:VCALENDAR