I am getting some great input from people. Thank you.
The method I was trying to think of is 'unfolding' which several people
identified, and Doug Carroll has directed me to multidimensional
unfolding.
Here is the problem that I may be faced with (given funding becomes
available)
:
1. Viruses contain sets of genes which for argument we will call A, B C,
D, etc.
2. We want to infer a relationship among virus 'species', though species
here is according to the virologists (or at least some of them) generally
not thought to apply in the strict evolutionary sense, so I use this term
loosely
3. The order of the genes within each virus species is different as a
result of significant genome rearrangements over time
e.g., ABCD ABDC .... CDBA ... etc
4. If we view the arrangement of genes among many viruses can we identify
families of viruses where the rearrangement is less within the family than
between families (sounds like clustering)
5. One approach is to define a metric on gene arrangement patterns (e.g.,
the minimum number of rearrangements needed to have identical gene
arrangement between two virus species) and proceed using standard
clustering
6. I also thought unfolding is something I need to consider, so thank you
for your input.
Anyone ever approach a problem like this?
Bill
---
Joint Meeting of the Interface and
Classification Society of North America
http://ilya.wustl.edu/if_csna_2005_meeting/
Abstracts and Registration Deadline is 4/9/05
William D. Shannon, Ph.D.
Associate Professor of Biostatistics in Medicine
Division of General Medical Sciences and Biostatistics
Washington University School of Medicine
Campus Box 8005, 660 S. Euclid
St. Louis, MO 63110
Phone: 314-454-8356
Fax: 314-454-5113
e-mail: [log in to unmask]
web page: http://ilya.wustl.edu/~shannon
On Wed, 9 Mar 2005, J. Douglas Carroll wrote:
> Maybe this is the usage of this term in some statistical fields, but the
> mathematical psychologist Clyde Coombs's (as far as I know original) use of
> the word "unfolding" implied finding a real valued continuum (a single
> dimension defined on at least an interval, not merely an ordinal, scale)
> such that all the input orders can be generated via a model in which each
> order is inversely related to the order of distances from an "ideal point"
> on that continuum. Unfolding was later generalized by some students of
> Coombs's to the multidimensional case, in which the single dimension was
> generalized to a multidimensional space, and the input orders were assumed
> to be inversely monotonically related to (Euclidean) distances from a set
> of ideal points in this multidimensional space. This generalization is
> referred to as "MULTIDIMENSIONAL unfolding".
>
> While it's certainly true that, in Coombs's original unidimensional version
> of unfolding analysis, an order can be associated with the unidimensional
> continuum resulting from unfolding analysis, its purpose was NOT to
> determine an ordering, but to determine an underlying
> continuum. Furthermore, the order defined by the resulting continuum will
> generally NOT be in any realistic sense a "consensus order"; in extreme
> cases its average rank order correlation (calculated by any reasonable rank
> order correlation coefficient) with the input orders could be zero, in fact.
>
> Doug Carroll
>
>
> At 10:27 AM 3/9/2005 -0600, shannon wrote:
> >Yes -- unfolding is the word. Thanks
> >
> >
> >Bill
> >---
> >
> > Joint Meeting of the Interface and
> > Classification Society of North America
> >
> > http://ilya.wustl.edu/if_csna_2005_meeting/
> > Abstracts and Registration Deadline is 4/9/05
> >
> >
> >William D. Shannon, Ph.D.
> >
> >Associate Professor of Biostatistics in Medicine
> >Division of General Medical Sciences and Biostatistics
> >
> >Washington University School of Medicine
> >Campus Box 8005, 660 S. Euclid
> >St. Louis, MO 63110
> >
> >Phone: 314-454-8356
> >Fax: 314-454-5113
> >e-mail: [log in to unmask]
> >web page: http://ilya.wustl.edu/~shannon
> >
> >
> >On Wed, 9 Mar 2005, Paul R Swank wrote:
> >
> > > Do you mean unfolding?
> > >
> > > Paul R. Swank, Ph.D.
> > > Professor, Developmental Pediatrics
> > > Medical School
> > > UT Health Science Center at Houston
> > >
> > >
> > > -----Original Message-----
> > > From: Classification, clustering, and phylogeny estimation
> > > [mailto:[log in to unmask]] On Behalf Of shannon
> > > Sent: Wednesday, March 09, 2005 9:19 AM
> > > To: [log in to unmask]
> > > Subject: statistical method
> > >
> > >
> > > What is the name of the statistical method which generates an order from a
> > > set of orders:
> > >
> > > Vote preferences: A > B > C > D
> > > B > A > C > D
> > > A > B > C > D
> > > A > C > D > B
> > > etc
> > >
> > > It is something like peeling?
> > >
> > >
> > > Bill
> > > ---
> > >
> > > Joint Meeting of the Interface and
> > > Classification Society of North America
> > >
> > > http://ilya.wustl.edu/if_csna_2005_meeting/
> > > Abstracts and Registration Deadline is 4/9/05
> > >
> > >
> > > William D. Shannon, Ph.D.
> > >
> > > Associate Professor of Biostatistics in Medicine
> > > Division of General Medical Sciences and Biostatistics
> > >
> > > Washington University School of Medicine
> > > Campus Box 8005, 660 S. Euclid
> > > St. Louis, MO 63110
> > >
> > > Phone: 314-454-8356
> > > Fax: 314-454-5113
> > > e-mail: [log in to unmask]
> > > web page: http://ilya.wustl.edu/~shannon
> > >
>
>
>
> ######################################################################
> # J. Douglas Carroll, Board of Governors Professor of Management and #
> #Psychology, Rutgers University, Graduate School of Management, #
> #Marketing Dept., MEC125, 111 Washington Street, Newark, New Jersey #
> #07102-3027. Tel.: (973) 353-5814, Fax: (973) 353-5376. #
> # Home: 14 Forest Drive, Warren, New Jersey 07059-5802. #
> # Home Phone: (908) 753-6441 or 753-1620, Home Fax: (908) 757-1086. #
> # E-mail: [log in to unmask] #
> ######################################################################
>
|
