eMOSS    |    EMBL Barcelona    |    Sharpe Laboratory    |    Webmaster

Userguide for staging mouse embryos

Purpose

Determine the precise age of a mouse embryo by means of its fore or himdlimb bud to characterize its development.
The present algorithm is trying to set a standard for the mouse limb growth based on a very large set of data collected at various stages, and it is capable to identify an embryo as belonging to a specific developmental age, with a typical time resolution in the range of one or two hours.
The range of stages spans from 10 days and 9 hours to 15 day and 21 hours post conception. If the forelimb is used instead, the range is limited to 10 days and 9 hours to 13 and 18 hours post conception.
The image uploading might not work properly with IE and old firefox browser versions.

Using the staging system

1. The staging system starts with a default dummy image. To stage your own image you need to upload it through the file dialog at the left of the page.
The embryo's limb image should be always taken frontally (dorsally or ventrally). To improve the visibility of the outline of the limb bud you can adjust brightness and contrast. The maximal filesize is 10 MB. Accepted formats are jpeg, png, bmp, and gif.

2. Draw points along the outline of the limb. Do not join the first and the last point. Points can be deleted in the inverse order of drawing by pressing x on your keyboard. You can choose to either identify single points or drawing a continuos line along the limb. You can always drag any points with the mouse. The image can be zoomed or dragged along the canvas.

3. To force the algorithm to stage the limb as left or right (looking from the dorsal side), click the corresponding check-button.
Default is "any": the algorithm itself will figure out the correct side.

4. No login or password are needed to use the staging system. If you insert a valid email you will be given an automatic link at the end of the staging process where your limb picture and the result of the staging will be uploaded, so that you can browse them and keep track of them. If you do not want this feature, leave the box blank.

5. Press "Stage Embryo". After a few seconds the output will be shown below the uploaded picture, which contains the relevant information about the staging of your mouse embryo, and will consist of a set of plots similar to the ones below.

The explaination of their meaning is the following:

- Plot on the left: the color points are the clicked points, the black line is the best-fitting shape that corresponds to your limb.
If you used the hindlimb, the red label on top-left shows the estimated stage of your mouse embryo as mE[day]:[hour] with uncertainty in units of hours.
If you used the forelimb you should read the second line (darker red, only available if the embryo is estimated younger than mE13:18).
The estimation of the number of somites is also given if it is available in that range.
- Top right: the $\chi^2_\nu$ value of the fit tells you about the reliability of the estimation:
values below ~5 are ok, values above ~10 indicate that the points do not fit well with any known shape.
For older ages a higher value can be expected because of the higher variability of the samples.
- Bottom right: the probability density as function of the limb age.
- Remember that you can right-click on the output image to save it for reference.

Troubleshooting and FAQ

Please report any problem or suggestion by email to: marco.musy at embl.es

- How should I interpret the errors?
All errors quoted correspond to one gaussian standard deviation, i.e. the limb age has 67% probability to lie within the calculated range. The asymmetric error of the median has the same meaning. The error can be large or small depending on how accurately the user has drawn the points and/or how accurately the limb matches one of the reference limb shapes.

- I get a warning message of "high" or "very high" chi-square.
A high chi-square value is due to the fact that the program is unable to find a good shape that fits the points because the user did not accurately draw them, and/or because there is genuinely no shape that resembles the user limb at this level of accuracy.
One should consider that for very high chi-square values the error might become meaningless. There is no general rule for the exceptions, so that the user should evaluate the situation case by case, e.g. judging if the black line in the plot on the left still corresponds to a reasonable fit to the points.
Bear in mind that if you have drawn a large number of points (e.g. >40) you are likely to get this kind of message because the fitting becomes precise enough to discriminate systematic differences between the limb you staged and the average limbs used in the database for comparison.

- I get a warning message of low number of points.
When too few points are drawn the errors can become large, and biases in the mean value of the limb age can be as large as the error or even larger. We suggest to identify at least ~30 points to obtain correct estimates.

- I get a warning message a length discrepancy.
If only draw a part of the limb which still fits an expected region in the database (e.g. you omit the flank/wrist), you can get this sort of warning. If the $\chi^2_\nu$ value of the fit is normal you can ignore this warning in most cases.

Bibliography

If you found this application useful, please quote this reference paper:
A Quantitative Method for Staging Mouse Limb Embryos based on Limb Morphometry
Marco Musy, Kevin Flaherty, Jelena Raspopovic, Alexandre Robert-Moreno, Joan T. Richtsmeier, James Sharpe
Development 2018 145: dev154856 doi: 10.1242/dev.154856 Published 5 April 2018

Other related papers:
The role of spatially controlled cell proliferation in limb bud morphogenesis.
A landmark-free morphometric staging system for the mouse limb bud.

COPYRIGHT 2018 ALL RIGHTS RESERVED - EMBL.ES