Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
Supporting Information Error Calculation: RMSE and Pearson Correlation Coefficient
 

Summary: 1
Supporting Information
Error Calculation: RMSE and Pearson Correlation Coefficient
The root mean square error (RMSE), as implemented in the Fakhouri et al. study,
was calculated using Giant protein concentrations as model inputs and the corresponding
lacZ mRNA concentrations as outputs [1]. Data points were taken at 20 different Giant
concentration levels, uniformly ranging from 0 to 1. Each term in the RMSE was
weighted by the number of actual data points over which each lacZ concentration was
averaged. In our sensitivity analysis, this weighted RMSE was used as the objective
function.
The Pearson correlation coefficient, as implemented in the Zinzen et al. study,
was calculated using Dorsal, Twist, and Snail protein concentrations as model inputs and
the corresponding rho and vnd mRNA concentrations as outputs, which can be found at
DVEx database (http://www.dvex.org) [2].
Enhancer-like structures used to calculate sensitivities of the Zinzen et al. model
For the results shown in Figure 3, we used two representative pairs of enhancer
structures proposed by Zinzen et al.. The remaining structures were analyzed as shown in
Figure S7. Each representative pair of enhancer structures is defined by the number of
modules and binding sites in each of the representative enhancers. An enhancer with
more than one module (Md > 1) is assumed to have identical, independent modules

  

Source: Arnosti, David N. - Department of Biochemistry and Molecular Biology, Michigan State University

 

Collections: Biology and Medicine