Detalle de: Crystallographic Methods

Our group pursues two distinct, but nevertheless, complementary kinds of scientific goals. On one hand the development of structural methods for the solution of macromolecular structures and on the other their application on projects of medical, biological or biotechnological relevance.

Crystallography allows us to see the main players in biological processes, even in functional complexes, providing conclusive evidence down to the atomic level. Nevertheless, this structural model cannot be directly calculated from the experimental data as we cannot measure the phases of the diffracted X-ray beams. Pushing the limits to solve this “phase problem” and establishing the experimental information beyond the model are the main aims of our research.

Our program ARCIMBOLDO, distributed by CCP4 and used worldwide, has allowed to phase ab Initio equal atom structures over twice as large and at resolutions of average, rather than exceptional experimental quality than previous methods. ARCIMBOLDO comes in three different flavours according to the user’s need, ARCIMBOLDO_LITE employs small and very accurate model fragments such as polyalanine α-helices and also has an specific mode for solving coiled coil structures, ARCIMBOLDO_BORGES employs libraries of local folds and ARCIMBOLDO_SHREDDER libraries of fragments from an homologous template or a deep learning prediction.

In our lab we also develop other programs for the determination and interpretation of crystallographic structures. ALEPH, a network-oriented approach for the generation of fragment-based libraries and for structure interpretation, ALIXE, a tool for combining partial solutions for fragment-based molecular replacement, SEQUENCE SLIDER, a new approach that models side chains on partial polypeptide traces in a multisolution framework, and the program package SHELX.

Our goal is to provide computational methods for macromolecular crystallography to better exploit the available experimental diffraction data and previous chemical knowledge resulting in the most effective, accurate, model-bias free and user-friendly structural determination, both in typical average cases and problem structures.