Skip to content

Latest commit

 

History

History
63 lines (34 loc) · 7.09 KB

File metadata and controls

63 lines (34 loc) · 7.09 KB

Getting started with geometric morphometrics analysis in the Dworkin lab

While this is a bit more than just a stub, much needs to be added.

Introductory readings

Geometric morphometrics requires some basic understanding of geometry, linear algebra and (for statistical analyses), statistics to get started. As with much we do in the lab, comfort in a scripting language that can deal easily with vectors and matrices (i.e., R or NumPy/Python) is a plus. No expertise is needed coming in, just some effort and patience to get comfortable with the material. If you want some good refreshers on the math side of things Khan Academy is always a good bet.

Geometric Morphometrics for Biologists - By Miriam Zelditch et al. We have a PDF of it in the lab dropbox space (so contact me for access to that space). We also have a physical copy that floats around the lab. This is a great introduction to the field, and has considerably depth to it.

Morphometrics with R - this is mostly useful once you are comfortable with the basics. When this was initially written, there was no single R library but a collection of functions. Much of which is also available in geomorph or shapes in R (see below). However, this book goes through a lot of the algorithms and basic functions and thus remains quite useful.

In the shared space on Dropbox we have notes from several Morphometrics workshops (both geared to beginners). One was run by Chris Klingenberg, and uses MorphoJ for running through exercises. The other was run by Dean Adams, Michael Collyer and Emma Sherratt and uses GeoMorph/R. As you are getting started it might be helpful to look at the same matetial in both (as they approach things from different perspectives.

How to use wing machine (tutorials)

From Will Pitchers

Software

StonyBrook morphometrics - This includes access to TPS software for morphometrics analysis, and links to many other useful resources. Get to know it!

geomorph/R. Geomorph has really really gone a great way to develop a very general purpose set of analysis tools for morphometrics. Their blog has lots of useful tutorials and tips and tricks. Indeed, much of what we had originally written in R is now somewhat redundant. Since geomorph has a larger user base than our lab scripts, it is more likely that bugs in the code are to be found. Thus I recommend getting used to it.

Morpho/R. Another R library for morphometrics. Some of the basiic tools overlap with geomorph, but there are many advanced plotting tools that differ. Also check out the blog associated with both Morpho and the authors other packages (i.e. Rvcg). The authors github page also may be useful.

mvMORPH/R. This is a relatively new R library for phylogenetic analysis using morphometrics data. I have not given it a whirl, but I am excited to try it out! The paper describing it is here. It is probably very useful to also get familiar with both ape/R and 'phytools'/'R' which are some of the standard toolbox for such analyses.

shapes/R. One of the older Geometric morphometrics libraries in R.

StereoMorph/R. This R library is for acquisition of 3D data from relatively inexpsensive set ups (i.e. not requiring a micro-CT scanner). The library was also co-written by Annat Haber (a BEACON post-doc who collaborated with me on a couple of projects). Check out the authors website for more details. Also here is the paper

MorphoJ. This is a GUI based set of tools written in Java. It is very easy to use and get started with. While it is currently not set up for high throughput analyses (like GWAS), or for complicated mixed models, it can do a suprising number of analyses. I also use this to teach geometric morphometrics to my Evolutionary Developmental Biology Undergraduate class. I even have a few tutorials up on youtube for this.

Wings & CPR. We use a pipeline utilizing some of the tools developed in the lab of our collaborator David Houle. In particular Wings to spline the wings in a semi-automated fashion (with human intervention to correct badly splined wings), and CPR (CP reader) to read in the CP files from wings, extract the landmarks and semi-landmarks, and perform the superimposition. One thing on my wish list (i.e. I wish I had time to code this), was to implement the tools in CPR in R (as opposed to a stand alone executible). This will enable easier setting up of the pipeline and we could use the superimposition algorithms as implemented in GeoMorph (and written in C++ which is called from R).

LORY. Another set of tools from the Houle lab (written by Eladio Marquez). Very useful for inferences of local shape changes (including nice visualizations).

Links to Dworkin lab .R scripts and functions for morphometrics

Ok, so I need to really update this (currently most of the development scripts are in the shared dropbox space). However, some of the older (and archived versions of our analysis pipelines can be found in the github repositories associated with our papers.

Pitchers et al. 2014

Sonnenschein et al 2015

Pitchers et al. 2013. This one is on DRYAD, but I should migrate it over soon.

Other useful resources

Morphometrics Library

morpho-tools

Evolutionary Morphing. Awesome visualizations!

Various databases

phenome 10k

facebase

Morphbank. I think this is the database that David Houle has put all of his Drosophila wings on.

GigaDB. This is where we hosted all of the images (I think ~6000) for a Drosophila wing dataset. All images, flat files with landmarks and scripts are available here. The scripts and landmarks are also on the DworkinLab github.