Machine learning for the geometrical identification of landmarks in shoot apical meristem images

Laboratory : Inria Team Mosaic, Reproduction & Développement des Plantes, ENS de Lyon
Duration : 6 months
Contact : Guillaume Cerutti (guillaume.cerutti AT ens-lyon.fr)
Keywords : Machine learning, 3D geometry, developmental biology, plant biology
Tools : Python, numpy, pandas, scikit-learn, jupyter

General Context

In flowering plants, most of the development takes place post-embryonically as the aerial organs (leaves and flowers) are successively formed over the lifetime of the plant. In a large number of species, these organs arrange into strikingly regular, often spiralling patterns. This patterning process called phyllotaxis can be traced back to the activity of a small niche of stem cells located at the tip of the growing stem : the Shoot Apical Meristem (SAM).

Using the model plant species Arabidopsis thaliana we study how these patterns are created by looking in detail at the formation of new organs. They actually emerge as primordia (groups of differentiating cells) in a very specific zone of the SAM, at the periphery of the undifferentiated central zone (CZ). The primordia appear with very regular timing and at spatial locations distant from a nearly constant angle. This strong spatio-temporal periodicity gives very interesting self-similarity properties to the system, and due to this, SAMs tend to present a high level of inter-individual similarity. The idea is to take advantage of this shape similarity to align different individual SAMs imaged using confocal microscopy onto a common reference frame. There, all individuals should superimpose almost perfectly, which makes it possible to perform population-scale statistics on quantitative measures of development.

See the following PDF for more details :

Offre de stage MOSAIC
Machine learning for the geometrical identification of landmarks in shoot apical meristem images

Voir en ligne : Galvan-Ampudia C., Cerutti G., Legrand J., Brunoud G., Martin-Arevalillo R., Azais R., Bayle V., Moussu S., Wenzl C., Jaillais Y.,Lohmann J., Godin C., Vernoux T. : Temporal integration of auxin information for the regulation of patterning, eLife, 2020.

Année : 2020-2021