Scientific interests of the AG Porto
The scientific interests of the AG Porto are mainly in statistical physics of complex systems in non-equilibrium, for instance as occurring in biology. In particular, we are concerned with physical modelling of diverse phenomena in molecular biology, molecular evolution, and structural bioinformatics, such as representation and organization of protein structures, evolution of protein sequences, or dynamical properties of complex networks. More `classical' fields of interest are structure and dynamics of soft matter as well as electronic and mechanical transport in mesoscopic systems. Some of our research highlights on these fields are given below.
Protein structure prediction
We have developed and studied a new approach based on structural profiles to efficiently identify near-native confirmations in ab initio protein structure prediction schemes such as ROSETTA, improving the quality of predicted structures (see publication R59 and related previous publication R35).
Protein folding dynamics
We have developed and studied a new model for the protein folding process based on structural profiles that yields a quite realistic folding dynamics despite the simplicity of the energy functional used (see publication R55 and related previous publications R35 and P15).
Protein structure representation
We have proposed a condensed vectorial representation of protein folds that evidences the relationship between protein structures and sequences (see publication R54 and related previous publications R35 and R42). The vectorial representations can be downloaded using SLOTH.
Protein structure alignment
We have developed a new efficient approach to detect similarities in protein folds based on structural profiles, yielding a distance measure in structure space. The approach is implemented in a tool called SABERTOOTH (see publication R45).
Reaction-diffusion processes on complex networks
We have studied the impact of topological correlation on the pattern formation and efficiency of the reaction-diffusion process A + B → 0 and have shown that disassortativity results in both enhanced pattern formation as well as increased reaction efficiency, which are usually mutually exclusive (see publication R50).
Prisoner's dilemma game on complex networks
We have studied the impact of topology on the dynamical organization of cooperation in the prisoner's dilemma game and have shown that assortativity and clustering can strongly facilitate cooperation even for a large temptation to defect (see publication R49).
Polymer adsorbed on disordered surface
We have studied the temperature-dependent behavior of a polymer adsorbed on a disordered surface using a simple model of a self-avoiding walk on a Sierpinski carpet and have shown how the polymer shape changes due to temperature (see publication R46 and related previous publication R41).
Neutral evolution of proteins
We have studied the neutral evolution of proteins to elucidate the statistics behind the process and have shown that the resulting neutral network in sequence space and the dynamics is rather complex (see publications R30 and P12).