Hello Plant Biologists of Ecolog,

A few months ago I sent out an email on ecolog to asses interest among the 
plant biologist community for an automated tool to phenotype
stomatal traits. I received a positive response from many of you andit’s clear 
the community needs a tool to make stomatal phenotyping
easier. Sven Eberhardt (Brown Uni.) and I have developed a tool to 
automatically count stomata and we have trained it on several hundred
Balsam Poplar epidermal micrographs. The method appears to work very well for 
Balsam Poplars (machine v. human annotation has r2 = 0.92).

It’s clear that the training set needs to grow, and that’s why I’m reaching out 
to the plant biologists of ecolog. I’m searching
for collaborators who can contribute several hundred high quality epidermal 
micrographs and then annotate them using the website we’ve
developed. Here are the parameters that your images need to abide by:\

+ The sample should come from populations within a species that encompasses a 
wide range of phenotypic variation.
+ Micrographs must be high quality, i.e. majority of image in focus, stomata 
clearly visible and unobstructed by epidermal features (e.g.
  trichomes). DIC images are the best.
+ Micrographs should be imaged at 20x or 40x.
+ Any preparation method is acceptable as long as it is a good prep.

I am particularly interested in finding contributors for the following groups: 
Pinus, Oryza, Arabidopsis, Brachypodium, Maize, Populus, Medicago,
and Mimulus.

Collaborators will need to contribute and annotate several hundred (~200- 500) 
images in the next few weeks. In recognition of that effort you
will be invited to join the project as a co-author, or acknowledged if you 
prefer. In addition to co-authorship/acknowledgement, contributors
will have access to a beta version of the tool/website, and the method will 
perform very well on your particular preparation methodology and
set of individuals.

If you are interested in learning more, please email me and include a zip file 
of a representative sample of images. If you need to
transfer large files, you can use UVM’s large file transfer service 
https://filetransfer.uvm.edu/.

Thanks for the interest Ecolog & I can’t wait to see what’s out there.

Sincerely,

Karl Fetter


Graduate Student
Keller Lab
Department of Plant Biology
University of Vermont
kfet...@uvm.edu

Reply via email to