Year: 2024
Type: article
Abstract: protein exists in, we compute a multi-state model from an X-ray diffraction pattern.We represent the system with multiple weighted structures (states) where the atoms of each structure can move independently of the other structures.We quantify the consistency of a multi-state model and the input information using a Bayesian scoring function that relies on the X-ray data as well as a molecular mechanics force field.We generate a sample of good-scoring multi-state models using a molecular dynamic-like sampling algorithm.We evaluate the performance of our method using both a synthetic and experimental benchmark.The key features of our approach are the following.First, we compute the set of structures that collectively best satisfy the data, rather than a set of single-state models that individually satisfy the data.Second, our Bayesian scoring function provides a conceptual framework for integrating the X-ray diffraction pattern with other sources of experimental information, including X-ray datasets of the system collected at distinct temperatures.Third, the method is implemented in our open-source IMP, relying on integration with Phenix, thus making it easily applicable to other systems.
Source: Biophysical Journal
Authors: Kibum Park, Eugene Shakhnovich
Institution: Harvard University
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: hybrid
Estimated APC paid: $2,500
OpenAlex ID: W4391657320