PIETRO DI LENA - UNIVERSITA' DEGLI STUDI DI BOLOGNA - PROTEIN INTER-RESIDUE CONTACT PREDICTION AND FOLDING

The protein-folding problem is considered one of the “grand challenges” in computer science and computational biology, with significant medical implications. From a purely computational point of view, the folding problem consists of predicting the three-dimensional (3D) structure of proteins from their one-dimensional (1D) aminoacidic sequences. The state-of-the-art of structure modelling methods and related sub-problems are assessed every two years at the CASP experiments. Starting from the first CASP edition in 1994, the protein folding problem has attracted the attention of a large community of researchers from di_erent fields, and important progress has been made in recent years.
In this talk, Prof. Di Lena will focus on inter-residue contact map representation of the protein structures and related research problems. Protein contact maps are square binary symmetry matrices that provide a two-dimensional (2D) rotation-invariant topological representation of the 3D protein structure. Protein contact map prediction is the CASP category that has registered the largest improvements in the most recent editions, due to the introduction of deep-learning approaches and novel statistical models based on Direct Coupling Analysis (DCA). In addition to the most relevant contact prediction approaches, he will discuss the application of predicted contacts to protein structure prediction and fold recognition.

• Poster (1 MB)