AutoSmooth – Signal Extraction Toolbox


AutoSmooth is a graphical interface for signal and derivative extraction from time series with nonstationary noise structure based on matricial Whittaker’s filter and information theoretic learning. The tool is applicable in the numerical decoupling of any systems of nonlinear differential equations into algebraic equations, and can be applied in reverse engineering of mechanistic model descriptions from multivariate experimental biologic time series.




The latest version of AutoSmooth application is now part of the Signal Extraction Toolbox, but it can still be downloaded separately from M.D.Anderson Cancer Center Bioinformatics Station.




The source code and stand alone application are provided at with free access and use, under a GNU GPL license. It can also be conveniently obtained as a module of the Bioinformatics Station resource ( ).




Stand Alone Installation – Windows


1-    Run MATLAB Component Runtime (MCR) installer:  (122 Mb).

2-    Download from

3-    Double click in Signal_Extraction_Toolbox.exe to start it.

4-    Important: MCR only works if you are connected to the internet.


Team Members

Marco Vilela

Graduate Assistant Student at the University of Texas MDAnderson Cancer Center


Dr. Jonas S Almeida
Professor of Bioinformatics at the University of Texas MDAnderson Cancer Center