The supplement request includes activities for FY 2010. The principal activities during this time to be supported by the supplement will be constructing tutorials and demonstration packages for the existing software we have developed. These activities are outlined in more detail below. These activities will accelerate the rate at which researchers new to these methods will be able to adopt them for their own projects.

Demonstrations and Tutorials The demonstration package will consist of an archive including the analysis plug-in, sample data files, support proce- dures, and html format documentation. The documentation will include embedded video tutorials. We plan to initially distribute video tutorials with embedded You-Tube videos on our group website. After the first few tutorials, we will evaluate this approach and determine whether it is adequate. If not, we will determine the degree to which the videos can be compressed for distribution along with the demonstration packages.

To generate the video tutorials, we will use video screen capture software to record the sequence of commands and actions required to perform the analysis. We will narrate the screen capture with instruc- tions, commentary, and rationalization for the various steps in the analysis procedure. In particular, we will point out where decisions and choices are made at different points in the analysis. This approach is modeled after the excellent video tutorials offered for the program SketchUp.

The purpose of the demonstrations and tutorials is to allow the users to see not only how to use the specific commands that we have implemented as plug-ins but also to see the logic in sequence of events and actions that need to be performed and are carried out the hidden Markov analysis. We will start with the analysis of single trajectories which is most applicable to long observations. We will include examples that assume a fixed number of states, that use the trajectory to determine the number of states, and that use a pseudo-continuous distribution of states. We will include examples for single-detector trajectories as well as multi-detector trajectories. We will include examples for intensity-only and intensity-plus-lifetime observations.

A second set of demonstrations and tutorials will be developed surrounding the integrated AS/GRIP HMM automated model initiation and multiple-molecule analysis. The approach in this set of demonstrations will be identical to the first except that in this case, we will be addressing the issues and steps of analysis associated with the automatic model initiation and the optimization of nuisance parameters.

We will prepare tutorials that address the three different approaches to global analysis that we’ve developed: concatenation, scaled concatenation, and Bayesian posterior likelihood analysis. The first method, concatenation, is the simplest. This method assumes that the nuisance parameters are identical for the different molecules. The second method, scaled concatenation, assumes that the nuisance parameters have been determined and that the interphoton times have been scaled to adjust for variability of excitation and detection efficiency. The third approach, Bayesian posterior likelihood analysis, evaluates the likelihood of the model as a function of its parameters for each molecule. The product of all Bayesian posterior-likelihood functions for all the molecules allows an effective global posterior likelihood function to be determined. The setup and execution of each of these methods is different and will require its own tutorial.