Seminar

Computational methods for carbohydrate structure determination, validation and analysis.

Prof. Jon Agirre

The introduction of intuitive graphical software has enabled structural biologists who are not experts in crystallography to build complete protein or nucleic acid models rapidly. In contrast, carbohydrates are in a completely different situation: scant automation exists, and users building models manually frequently trip over legacy issues such as incorrect dictionaries or non-standard atom naming, which evidence a historical lack of methodological support for carbohydrates. Sugars are stereochemically complex and, as pyranose rings, have clear conformational preferences. And despite this, all refinement programs may produce high-energy conformations at medium to low resolution, without any support from the electron density; this problem renders the affected structures unusable in glyco-chemical terms. Bringing structural glycobiology up to ‘protein standards’ is thus requiring a total methodological overhaul. Time is of the essence, as the community is steadily increasing the production rate of glycoproteins, and electron cryo-microscopy has just started to image them in precisely that resolution range where crystallographic methods falter most. In this talk, he will introduce our latest methodological developments, designed to streamline and automate hitherto error-prone processes, effectively aiding crystallographers and electron microscopists alike in producing correct atomic models with confidence.