SRS_Synthetic: ------------- M. Serkan Apaydin Send questions/bug reports to: apaydin@cs.stanford.edu 07-17-03 code also contributed by David Hsu. Description: ------------ Code to construct a roadmap in a 2-D synthetic energy landscape, run Monte Carlo simulation, and compute pfold on this roadmap. Also included is a set of matlab files that can be used to *) compute pfold and plot its comparison with those obtained from Monte Carlo simulation *) plot the landscape, as well as visualize the roadmap on this landscape. Absolutely no warranty. Used outside packages: 1) ANN : for nearest neighbor computations 2) meschach : for linear algebra computations Installation instructions: -------------------------- 1) tar xvfz srs_synthetic.tgz (which you should already have done) 2) cd srs_synthetic (ditto) 3) make compiling/building ANN: (if included library does not work) ----------------------- If there is a problem in linking due to some functions provided by the ANN package, go to ANN directory (ann_0.2), type: make realclean make ANN will give you different options to prepare the ANN library for different platforms, choose the one that fits yours (if it is linux, make sunos5-g++ works in my linux machine). Then go back to srs_synthetic directory and type make again. compiling/building meschach: (if included library does not work) ---------------------------- Similarly, if the included meschach library (meschach/libmeschach.a) does not fit your system, you will have to rebuild it. To do this, go to meschach directory, and type: make realclean ./configure --with-sparse make sparse make clean Usage: ------ ./predict Index is optional (0 by default). This creates a roadmap, and computes pfold. Writes the result to pfold%d_%d.txt, numNodes, index. Some files need be in the current dir from which predict is run. These are: 1)pfold100MC.txt : The roadmap constructed includes some nodes read from this file. 2)parameters.txt : The parameters of the synthetic landscape are read from this file. The documentation for the code is generated with doc++ and is available in the html directory under the src directory. By commenting out the functions inside main() in predict.C, one can also have a Monte Carlo Simulation starting from a given node of the roadmap. Have fun!