----- Forwarded message from Chris
Klingenberg
Date: Sat, 18 Aug 2012
03:29:14 -0400
From: Chris Klingenberg
Reply-To: Chris Klingenberg
Subject: Re:
help
To: [email protected]
Yes, you can use centroid size as a measure of size for calculating
asymmetry of size etc. You would not want to normalize, because that
would eliminate the possible differences that you're after.
You use one Procrustes fit for all wings and then sort them out by
fly and side.
Best wishes,
Chris
----- Forwarded message from
Alfonso Caetta
Date: Tue, 14 Aug
2012 13:41:23 -0400
From: Alfonso Caetta
Reply-To: Alfonso Caetta
Subject:
help
To: [email protected]
----- End forwarded message
-----
--
***********************************
Christian Peter Klingenberg
Faculty of Life Sciences
University of Manchester
Michael Smith Building
Oxford Road
Manchester M13 9PT
United Kingdom
E-mail: [email protected]
Phone: +44 161 2753899
Skype: chris_klingenberg
***********************************
----- End forwarded message
-----
Hi Alfonso
On 18/08/2012 05:30,
[email protected]
wrote:
Hi,
we are trying to use landmarks to
analyze the effect of some genetic and environmental variables
on the
developmental instability of the fly wing. We are using
MorphoJ with landmarks
obtained using TPSdig.
We want to measure size
variation between genotypes/temps (using centroid size
variation) and shape
variation (using procrustes fit variation). We wanted to use
both fluctuating
asymmetry, and total within-group variation, as our measures
of developmental
instability.
As newcomers to this field, we have a
couple of basic questions:
1) Can raw centroid size
difference (R minus L) be used to gauge wing size variation,
or must these
values be corrected somehow to account for allometry or
overall size differences
that may exist between genotypes? i.e., how do we normalize
the R-L values so
that they are comparable across genotypes and
treatments?
2) When doing procrustes fit to estimate
fluctuating asymmetry values, must we perform an individual
procrustes fit for
the two wings of EACH individual fly, followed by an
estimation of procrustes
distance between the two wings' landmark profiles? Or,
alternatively, can we
perform a single procrustes fit for ALL wings of each genotype
say, and then
using an L vs R classifier in MorphoJ, estimate procrustes
distances between
wing pairs?
I hope these questions make
sense!!
thanks
--
Alfonso
Caetta
Columbia University, NY
