Automatic spike train analysis and report generation. An implementation with R, R2HTML and STAR.

Multi-electrode arrays (MEA) allow experimentalists to record extracellularly from many neurons simultaneously for long durations. They therefore often require that the data analyst spends a considerable amount of time first sorting the spikes, then doing again and again the same basic analysis on the different spike trains isolated from the raw data. This spike train analysis also often generates a considerable amount of figures, mainly diagnostic plots, that need to be stored (and/or printed) and organized for efficient subsequent use. The analysis of our data recorded from the first olfactory relay of an insect, the cockroach Periplaneta americana, has led us to settle on such "routine" spike train analysis procedures: one applied to spontaneous activity recordings, the other used with recordings where an olfactory stimulation was repetitively applied. We have developed a group of functions implementing a mixture of common and original procedures and producing graphical or numerical outputs. These functions can be run in batch mode and do moreover produce an organized report of their results in an HTML file. A R package: Spike Train Analysis with R (STAR) makes these functions readily available to the neurophysiologists community. Like R, STAR is open source and free. We believe that our basic analysis procedures are of general interest but they can also be very easily modified to suit user specific needs.