Structural Bioinformatics - Database and prediction methods

Bioinformatics is known as the discipline of exploiting, interpreting and annotating biological data using algorithms, i.e. computational methods. The data most frequently exploited are those relating to genome and protein sequences. However, while knowledge of this information is a prerequisite for characterizing a biological macromolecule, it is not sufficient to understand its function. The latter depends on the appropriate spatial arrangement of the constituent units of these biological macromolecules, such as nucleotides and amino acids. Knowing and analyzing the 3D structure of macromolecules is therefore a necessary step towards a better understanding of their function, and therefore their possible dysfunctions. Various biophysical techniques provide access to this structural information. Circular dichroism, for example, can provide information on the structure of the polypeptide skeleton of proteins, or on the shapes adopted by DNA sequences. Crystallography coupled with X-ray diffraction and nuclear magnetic resonance can provide information on structure on an atomic scale. Exploiting these data, designed to shed light on the characteristic properties of macromolecules, requires the use of computational methods known as "structural bioinformatics" or "molecular modeling".

By analyzing the information available on the 3D structures of proteins, it has been possible to demonstrate close relationships between sequence and structure, in particular the conservation of 3D structure for similar sequences, but also for very different sequences, especially those derived from a common ancestor (homologous sequences). This property can thus be used to predict the 3D structure of a macromolecule from its sequence. This is a major challenge for the projects that have followed on from the massive sequencing projects. These include initiatives to set up and automate procedures for obtaining the 3D structure of proteins. (Protein Structure Initiative) . Similarly, the impact of mutations at distant positions along the sequence can be better understood, as the 3D structure can spatially group these positions. Knowing the structure also enables us to better understand the fine interactions at play between partners, which depend on their relative positions. This makes it possible to understand and improve ligand specificity, and even to potentially design new proteins. (protein design) or new drugs, a major challenge for the coming decades.

Finally, a biological macromolecule is not a rigid object, but rather one capable of adapting to different environments and different partners in order to accomplish its function. This adaptation, which takes place over time, can induce relatively limited or, on the...