Jan. 18, 2018 added by Alain Kamgoué
How can we extract the 3D rDNA structure in living cell from a 3D map of fluorescent microscopy. To begin, we convert the fluorescent microscopy 3D output into MRC format. This convertion gives us the oppurtinity to use Segger. Segger is an extension of UCSF Chimera using 3d watershed segmentation. This segmentation allows us to obtain different regions with their local maxima. At this stage, we have two possibilities to build the first support points of our reconstruction. For each region, we can choose the local extrema as the first support point; the other option is to choose the centers of mass of each region as the first support points of our reconstruction. Note that for very homogeneous structures, the two possibilities merge. At this stage, the 3d reconstruction of the rDNA can be done naïve way by connecting the points obtained previously. This approximation is also called first level approximation. In order to improve our reconstruction, we have introduced secondary points. For two adjacent regions, the two points that minimize the distance between the two regions are chosen. By adding these points in our reconstruction, we are talking about second-level reconstruction. All connection points obtained, the next question is that of their connection. For this, we will consider the graph whose set of nodes consists of all the points defined above. By traversing the graph in depth, we build all the possible paths. Of all these paths, which one approaches our 3d structure? To answer this question, we calculated the lengths of all paths. The shortest path is considered the one that best approaches the 3d structure.