I have extracted the bitmap of the combined field (part a of Fig 6), and removed the dendrites from around each cell into a new image. (I extracted the figure from the pdf file, using pdfimages tool, as this gave a much higher resolution image of the field.) Custom-built software then found the centre of mass and number of pixels for each object. This information is stored in w81_pos.txt.safe
This figure overlaps the bitmap with circles centred on the cells with equivalent area to the cells. The fit is good. Green is for on cells, red for off cells.
I then converted the coordinates from pixels to microns (um) using the fact that the scale bar in the figure is 250um, i.e. the width of the field is 750 um.
This produced three files where the coordinates are stored (units are microns):
On cells - w81s1on.txt
Off cells - w81s1of.txt
Sample window (xlo xhi ylo yhi) - w81s1w.txt sample window
If the circles accurately denote the somas, then the pixel counts indicate the somas have an mean radius of 9 um with a range from 7 to 13 um. These values can only be approximate.
The R script w81s1_check.R creates the
following plots to check that the points had been created ok.
Plot of the cells (assuming 9um radius): field.
Histograms from figure 6.