Recent Publications

Image 1

Fig. (2). In the above figure we have the connection between an element (CDK family) of the protein space and a sub-space of the chemical space (CDK inhibitors for which crystallographic structures are available). This connection is mediated by the scoring function capable of predicting the ligand-binding affinity. In the above scoring function space, the α's indicate the relative weight of the explanatory variables (x's) and f indicates the response variable to predict ligand-binding affinity. 


Logo of TabaGithub Logomark


Laboratory of Computational Systems Biology

Research Highlights

Welcome to the website of the Computational Systems Biology research group of Prof. Walter F. de Azevedo Jr.

We are a research group working on the development computer models to simulate natural systems. Our research interests are interdisciplinary with following major emphases:   


Our main goal is the development of research projects focused on Computational Systems Biology. We have been working on the development of computational models for unraveling the molecular mechanisms underlying enzyme inhibition and protein-ligand interactions. These computational models can be used to predict binding affinity of a potential inhibitor for an enzyme, such knowledge has the potential to speed up drug discovery and decrease the cost of development of new drugs. Furthermore, the availability of computational models to predict binding affinity based on the atomic coordinates of protein-ligand complexes adds flexibility to the process of drug discovery, since it allows us to computationally test different scenarios where a potential new drug may interact with a protein target. With such methodologies, we adopt the principles described in the seminal work of Stephen WolframA New Kind of Science, where computer simulations of complex systems play a pivotal role to bring deeper insights into the functioning of nature.    


Exploring the Scoring Function Space

We envisage protein-ligand interaction as a result of the intersection between the protein sequence space and the chemical space , and we propose to approach these sets as a unique complex system , where the application of  computational methodologies  could contribute to establishing the physical principles to understand the structural basis for the specificity of ligands for proteins. Such approaches have the potential to create novel semi-empirical force fields to predict binding affinity with superior predictive power when compared with classical methodologies. We propose to use the abstraction of a mathematical space composed of infinite computational models to predict ligand-binding affinity, named here as scoring function space . By the use of supervised machine-learning techniques , we can explore this scoring function space to build a computational model targeted to a specific biological system, for instance HIV-1 Protease . We developed the programs  SAnDReS  and Taba to build computational models to predict ligand-binding affinity.  SAnDReS  and  Taba  are integrated computational tools to explore the scoring function space.        


A view of the scoring function space as a way to develop a computational model to predict ligand-binding affinity.

Here you find overall information about our research projects , publications , tutorials , and teaching material.    


Research Projects

The availability of more than 120,000 experimentally determined macromolecule structures paved the way to investigate computer models to predict three-dimensional structures of proteins and to simulate protein-ligand interactions. Furthermore, integrating protein structures and binding affinity information creates a favorable environment for the development specific scoring functions to predict ligand-binding affinity.  We are working in the development of computational tools to analyze protein-ligand interactions to carry out protein-ligand docking simulations and to predict three-dimensional structures of proteins.         

Re-dock result for the structure of cyclin-dependent kinase 2A4L . Scatter plot and ROC Curve generated by SAnDReS.

A small sample of our research projects running right now at      

   - SAnDReS   PubMed   
   - Taba: A Tool to Analyze the  Binding Affinitty    PubMed
   -Fine Tuning of Scoring Functions    PubMed  
   -Application of Molecular Docking to Nanobiotechnology    PubMed    
   -Protein-Ligand Docking Simulations    PubMed   
   -Machine Learning Algorithms Applied to Systems Biology    PubMed     
   -Machine Learning for Development of Scoring Functions    PubMed    
   -Development of Scoring Functions for CDK    PubMed   
   -Machine Learning Algorithms Applied to Bioinformatics    PubMed      




Contact Us

If you have any question, please feel free to contact Prof. Walter F. de Azevedo Jr. e-mail:

Join Us

If you are interested in computational systems biology please contact us (e-mail:

Powered by Python Powered by NumPy Powered by SciPy Powered by Pyzo Powered by scikit-learn Powered by MathJax