BEGIN:VCALENDAR
PRODID:-//Columba Systems Ltd//NONSGML CPNG/SpringViewer/ICal Output/3.3-
M3//EN
VERSION:2.0
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
DTSTAMP:20201005T081640Z
DTSTART:20201014T130000Z
DTEND:20201014T140000Z
SUMMARY:Merle Behr - Learning Compositional Structures
UID:{http://www.columbasystems.com/customers/uom/gpp/eventid/}u12l-kfw9e7
n0-q61whq
DESCRIPTION:(via Zoom)\nJoin us for this seminar by Merle Behr (UC Berkel
ey) as part of the North West Seminar Series in Mathematical Biology and
Data Sciences. Details of the full series can be found here https://www
.cms.livjm.ac.uk/APMSeminar/\n\nAbstract: Many data problems\, in partic
ular in biogenetics\, often come with a highly complex underlying struct
ure. This often makes it difficult to extract interpretable information.
In this talk we want to demonstrate that often these complex structures
are well approximated by a composition of a few simple parts\, which pr
ovides very descriptive insights into the underlying data generating pro
cess. We demonstrate this with two examples.\n\nIn the first example\, t
he single components are finite alphabet vectors (e.g.\, binary componen
ts)\, which encode some discrete information. For instance\, in genetics
a binary vector of length n can encode whether or not a mutation (e.g.\
, a SNP) is present at location i = 1\,…\,n in the genome. On the popula
tion level studying genetic variations is often highly complex\, as vari
ous groups of mutations are present simultaneously. However\, in many se
ttings a population might be well approximated by a composition of a few
dominant groups. Examples are Evolve and Resequence experiments where t
he outer supply of genetic variation is limited and thus\, over time\, o
nly a few haplotypes survive. Similarly\, in a cancer tumor\, often only
a few competing groups of cancer cells (clones) come out on top. \n\nIn
the second example\, the single components relate to separate branches
of a tree structure. Tree structures\, showing hierarchical relationship
s between samples\, are ubiquitous in genomic and biomedical sciences. A
common question in many studies is whether there is an association betw
een a response variable and the latent group structure represented by th
e tree. Such a relation can be highly complex\, in general. However\, of
ten it is well approximated by a simple composition of relations associa
ted with a few branches of the tree. \n\nFor both of these examples we f
irst study theoretical aspects of the underlying compositional structure
\, such as identifiability of single components and optimal statistical
procedures under probabilistic data models. Based on this\, we find insi
ghts into practical aspects of the problem\, namely how to actually reco
ver such components from data.\n\nDetails of the zoom meeting ID and lin
k will be sent via the Maths in the Life Sciences mailing list. If you h
ave missed the e-mail please contact the seminar organiser directly (car
l.whitfield@manchester.ac.uk or i.siekmann@ljmu.ac.uk).\n\nTo subscribe
to the mailing list for this event series\, please send an e-mail with t
he phrase “subscribe math-lifesci-seminar” in the message body to listse
rv@listserv.manchester.ac.uk
STATUS:TENTATIVE
TRANSP:TRANSPARENT
CLASS:PUBLIC
END:VEVENT
END:VCALENDAR