Proteins are often organized in large, dynamic molecular machines. A mechanistic, atomic-resolution understanding of these machines would provide fundamental insights into the clockwork of the complex machinery underlying human life. However, machines such as the 26S-proteasome have resisted crystallization attempts for decades.
Cryo-electron microscopy (cryo-EM) has provided insightful maps of molecular machines, but these lack atomic resolution. In principle, molecular machines can be assembled from atomic homology models of the subunits. However, homology models often contain large errors, causing existing assembly methods to fail.
Here I present a multidisciplinary approach for the assembly of large molecular machines from cryo-EM maps. I will combine computational techniques from cryo-EM, homology modelling and interface prediction into ATTRACT, the only docking method that integrates subunit flexibility into the initial search. Driven by a cryo-EM density map, ATTRACT’s coordinated motions will reduce the initial homology model errors, bending the subunits into the correct shape while they are being assembled.
The challenging nature of the problem requires the integration of approaches. After firm assessment of the abilities of the method, it will be applied to experimental density maps of the proteasome and other molecular machines, providing new fundamental insights when neither complex nor subunits are known.
Convocatoria de propuestas
Consulte otros proyectos de esta convocatoria