Custom software

Quick Guideto AT synapse analysisusing Image AnalysisSuite

Note: It is very important to start with perfectly aligned images within volumes and across sessions. Thus,it is recommended that all images are aligned and dewarped.

1. (We are starting from perfectly aligned image stacks – in virtual stack format, e.g. a series of images) use the function: find_centroid_smart. This function will find all of the centroids for each protein punctum in the image stack and output each in a .csv file (example syntax: details of the different parameters are written in the function, use either: help find_centroid_smart at the matlab prompt or open the m-file) 
   find_centroid_smart('type',2,'type2',[0,1,5],'object',1,'threshold',[0.03:0.01:0.15],'norm',[0.9999],'zfilter',0);
2. Move all of the calculated centroids for all of the channels into a single folder. If they are not already in a single folder.
3. Next run punc_locodist2. This function will prompt you to open the protein channels (CSV files from find_centroid) you desire to analyze and generate a matrix of distances held in the two variables punc_data and all_data [in the example]. Note: if voxel is isometric leave the inputs to the function blank. The z definition in the example of 0.7 is only because our normal voxel is 1,1,0.7. (example syntax: again details of parameters in file) [punc_data,all_data] = punc_locodist2('z',0.7);
4. Run savpivots_all. This function will generate pre-postpairs that is central to synapse generation. The input to savpivots is determinate. It’ll take the punc_data and all_data variable from the previous step. Ori is the filename of the protein you want to be the origin (post or pre), termi is what you want to be the opposing point of the origin (post if pre, pre if post). Ch3 is a third channel that is used to define a synaptic pair. A putative ori-termipair is only valid if thereis a Ch3 point within the distance set by prepost_dist. Filter3 if set on is an algorithm that attempts to reach the most parsimonious allocation of the Ch3 points so that no ori-termi pair shares the same Ch3 point.For more info on prepost_dist and filter 3 setting see the help text in punc_colo_all. Multi sets multi threading for rotations on or off. Path is the location of the files, and rotation is necessary for SubSynMap however it greatly increases runtime. (examplesyntax: again detailsof parameters in file)
   savpivots_all(punc_data,all_data,ori,termi,ch3,prepost_dist,filter3,multi,path,rotation);
5. At this point you have putative synapses classified, and they should be saved out to its own directory in .csv format. If you want to calculate density, I suggested using fillstat_vstk, and calculate the volume to be excluded due to the nucleus.If you want to attempt SubSynMap then make sure the rotate parameter is turned on, and the files should already be saved in the same directories. Look for files with heading 3dvert and flat. 3dvert are vertices of that particular channel rotated so that the ori-termi axis is on a single cartesian axis. These vertices are in 3D x,y,z space, which is difficult to visualize, flat is the same verticies flatted into 2D, so that the 3^rd axis is removed through rotation. You could visualize the flat points as a raster plot or use dhist and dhist3.

Category