PREMER (Parallel Reverse Engineering with Mutual information & Entropy Reduction) is an open-source, multi-platform software tool for inferring network structures from data, using information-theoretic measures. Many systems (biological or otherwise) can be meaningfully represented as networks, that is, as a set of nodes (variables) connected by links (interactions). 

PREMER predicts the existence of network links, estimates their relative strength and direction, and provides a visual representation of the inferred system. It also features a data preprocessing step which enables the use of datasets with missing values and/or outliers. Furthermore, it allows the inclusion of previous knowledge: the user can specify if certain reactions are known to be nonexistent.

PREMER is a general purpose tool, which has been developed with biological networks in mind but can also be applied in other areas.

Implementation and Requirements

PREMER is available in two implementations: for MATLAB/Octave, and for Python. 

PREMER runs on Windows, Linux and OSX. It only requires a MATLAB, Octave, or Python environment to run in each of these operating systems.

PREMER's core computations are performed in compiled code written in FORTRAN F90, which allows for more efficient computations than Matlab, Octave, or Python. It has been implemented using OpenMP directives, which enable it to run seamlessly in parallel environments, thus allowing for additional speedups in performance.

PREMER is released under the free and open source GNU GPLv3.


  • The first publicly available version of PREMER (v1.0) was released in March 2015.
    • PREMER v1.1 was released in November 2015.
  • PREMER v2.0 was released in March 2017.
    • PREMER v2.1 was released in October 2017.


This software is distributed under