AmPeps at the DEB2021 course

This page explains what you can do in preparation of an entry before the course DEB2021 starts and what to do next. Please remember that we see an entry in the AmP collection as a publication, meaning that
  • submission means that you agree that it will be posted on the AmP website (after curation). If it contains data outside the public domain, please ensure that all involved parties agree.
  • you can invite others to help data-collection/writing. It is generally a good idea to include people who contributed to the data in the author-list of the entry; entries have an optional field aknowledgments, where you can e.g. mention project numbers and supporting funding.
  • the bibitem of the corresponding bibkey of data-sets should make clear where the data came from. If you measured the data yourself, please mention this in a bibitem of type "mics".
  • your entry will be carefully curated, meaning that what will be posted might differ from what you submit. There will, obviously, be feedback.
  • please make use of the comment fields as much as possible, to give some background for the data. If you selected a value within in a range, please mention the range in the comment. Different sources typically give different values; mention this. If you guessed the temperature, please mention that explicitly. Does the data apply to females or both sexes? If it only applies to males, the code-name has a trailing "m"; female is the default gender in AmP. If data is length, make explicit how length is defined. Different length-measures require different shape coefficients.
  • if males differ from females, please include a discussion point that specifies in what parameters they are supposed to differ. If you excluded particular data, via zero weights, motivate this in a discussion point.
  • if facts exist that are worth knowing, especially those that have implications for the interpretation of data, please mention this in one or more items in the optional section facts, and provide references. You can also think of preferred temperature, to support your choice for "Typical body temperature", or supporting facts for eco-codes, or empirical length-weight relationships if you had to make use of this.

We assume that you have seen the videos

  • DEBsite on the setup of the AmP web site.
  • DEBdmod on DEB models and parameters in AmP
  • DEBbody on body size as an emergent trait of metabolism.
  • DEBfood on food correction.
  • DEBtemp on temperature correction

What to do next depends if your species is already in the AmP collection or not. If your pet is a particular geographical variety of an existing species that might deviate in parameter values, it counts as a new species. In that case you extend the name of the species with a code for the geografic area, see e.g. Poecilia reticulata in the AmP collection.

Your pet is not already in AmP

First collect data with references for your species. It should at least have
  • size at birth, puberty, and ultimate size. If size is some length, you should hunt for at least one value to couple length to weight.
  • reproduction as ultimate rate, or fecundity at a known age. Better is fecundity/reproduction as function of length or weight.
  • maximum life span.
  • size (= length and/or weight) as function of time (preferably time since birth), if not available: age at puberty.
All rates and times require a temperature; the more data the better. See the AmP collection for inspiration.

Watch the videos

  • DEBecoc on the use of eco-codes.
  • DEBampe on the AmP entry prepare system.
and read the manual for the AmP entry prepare system carefully. Please mail us if something is not clear.

Once you have produced the 4 source files with AmPeps, you can always add more data by editing the mydata-, pars_init- and predict-files.

We have the experience that it sometimes can happen that the procedure fails, for reasons that are not fully understood. Retrying, by typing AmPeps, then hit resume and pause/save and quit AmPgui, continue with AmPeps, the problem disappeared in cases that we tried. In the unfortunate case that AmPeps continues to fail, please report this to Bas Kooijman and append the results_my_pet_backup.mat file, if it was produced. You can always download-and-edit empty templates for the source files and try DEB species explorer to find useful initial parameter values for the pars_init file.

Your pet is already in AmP

We assume here that you want to extend/improve an existing entry. The general rule is that all existing data in an entry remains in the update, unless arguments exist to remove it. These arguments should be given in one or more discussion points, as well as a (brief) explanation for the update (what changed and why).

The best way to proceed is to edit the mydata-, pars_init- and predict-files directly (in Matlab), using other entries as example. If you try AmPeps, it will discover that the species is already in the collection as soon as you enter its name (without typos) and proceeds by loading the 4 source files in the Matlab editor, skipping all other edits; a fast way of copying files. Notice that species frequently change names nowadays. AmP follows the nomenclature of species that are accepted in the Catalog of Life. The species_list gives an overview on what entries have what data, allowing you to select example-entries. Use the search-fascility of your browser to locate example-species, but watch the model type. DEBwiki describes the codes that are used for zero- and uni-variate data.

Especially if your new data concerns varying food and/or temperature conditions, watch the video

  • DEBvpar on varying food and/or temperatures.
Food and/or temperature trajectories are frequently specified as knots of a spline (using interpolation). Entries that have such constructions can be identified by searching for the string spline with function select_predict. Make sure that your example-species has the same DEB model. Likewise, seasonally fluctuating temperatures are frequently modelled with a sine or cosine function. Identify example entries by searching for the strings sin or cos in the predict files.

Once you have the 4 source files during the course

The next step is to estimate parameters; the course crew will help you with this. Please watch the videos
  • DEBcpar on parameter estimation in practice.
  • DEBppar on point parameter estimation.

If you arrived at a reasonable fit, the next crucial step is to find out if the parameter values make biological sense. See the video

  • DEBrpar on comparison with reference species

The final phase of this course exercise is the most creative and important one: to interpret the results in terms of the biology of your species. Find traits that you did not expect. Is this new biological insight or an invitation to reconsider data and re-estimate? Focus on this part in your lecture on your entry.

After the course

After the course, you can do a more comprehensive in-depth study on the traits of your species, using the tools as provided by AmPtool. See the videos
  • DEBampt on AmPtool/AmPdata,
  • DEBshst on the use of function shstat for plotting
  • DEBpatt on patterns in parameter values
The general idea is that your entry replaces the electronic material what you would otherwise append to a paper on the comparative traits of your species. This material was carefully reviewed, which hopefully helps in getting your paper accepted. All methods that were used, including the models, are published, so a few references will do, which allows you to fully focus on the biology or applications. We hope that the environment of the course and the symposium helps you with identifying promising candidates for productive collaboration.

You can always submit revisions of your entry.