ИСТИНА

The cryo-EM revolution is now driving the progress in our understanding of chromatin structure by solving the structures of nucleosomes and their complexes. The number of structures in the PDB database has doubled over the last three years and is rapidly growing. However, harnessing helpful information from this wealth of structural 3D data is challenging. Variations in components’ naming nomenclature, chain identifiers, amino acids, and nucleotide numbering and lack of readily available tools for systematic structure analysis and comparison adds to the complexity of the problem. We propose an automatic python toolkit and a web instrument as an easy-to-use online web-based solution to this problem. In our analysis approach, all nucleosome-containing structures are automatically identified in the PDB and classified according to their content and spatial organization. Secondly, all structures are brought to a common spatial reference frame by identifying their symmetry and superhelical axes. This common reference frame provides new ways to explore and compare the structures. Thirdly, identification of related DNA positions allows for comparing DNA geometry, twisting, and interactions with histones across the diverse set of structures. Currently, we were able to extract, annotate and analyze 353 structures automatically. We discovered 143 structures of complexes of nucleosomes with various proteins, which we classified by their type and associated function. Such bulk analysis allows us to create an extensive map of molecular interactions with nucleosomes. After alignment to standard reference frames, we were able to find several discrepancies in DNA positioning on nucleosomes which may be an indication of DNA sliding or the result of the suboptimal fitting of DNA into poorly defined EM density maps.