PDBe: improved accessibility of macromolecular structure data from PDB and EMDB

Abstract

The Protein Data Bank in Europe (http://pdbe.org) accepts and annotates depositions of macromolecular structure data in the PDB and EMDB archives and enriches, integrates and disseminates structural information in a variety of ways. The PDBe website has been redesigned based on an analysis of user requirements, and now offers intuitive access to improved and value-added macromolecular structure information. Unique value-added information includes lists of reviews and research articles that cite or mention PDB entries as well as access to figures and legends from full-text open-access publications that describe PDB entries. A powerful new query system not only shows all the PDB entries that match a given query, but also shows the 'best structures' for a given macromolecule, ligand complex or sequence family using data-quality information from the wwPDB validation reports. A PDBe RESTful API has been developed to provide unified access to macromolecular structure data available in the PDB and EMDB archives as well as value-added annotations, e.g. regarding structure quality and up-to-date cross-reference information from the SIFTS resource. Taken together, these new developments facilitate unified access to macromolecular structure data in an intuitive way for non-expert users and support expert users in analysing macromolecular structure data.

Figure 1.

An example of an entry summary page showing the organisation into three main areas (highlighted with dashed lines). The top panel shows the essential details and access to a picture gallery. The right-hand panel contains ‘Quick links’ to pages with more detailed information, file downloads and 3D viewer. It is also used to provide other relevant information, e.g. if the entry has been cited or mentioned in reviews or other articles. The main body of the page is divided into four sections as described in the text, providing summary information and links to the detailed pages for each of the sections.

Figure 2.

(A) Image highlighting the location of the six copies of globin d (blue) within the giant haemoglobin complex from Glossoscolex paulistus. PDB entry 4u8u. (B) The human rhinovirus capsid (PDB entry 4rhv) contains 60 copies each of four different proteins: VP1 (green), VP2 (yellow), VP3 (blue) and VP4 (cyan). (C) The location of the eight copies of Pfam domain ‘Pumilio-family RNA binding repeat’ is highlighted in this image of Human Pumilio 1 protein, PDB entry 3bsx.

Figure 3.

Interactive visualisation tools displaying annotations for each individual protein molecule. The figure shows the Pfam domain (PF00198: 2-oxoacid dehydrogenases acyltransferase (catalytic domain)) in protein molecule 'Dihydrolipoyllysine-residue acetyltransferase component of pyruvate dehydrogenase complex' (PDB entry 1dpb) highlighted in the sequence feature view (1D), topology diagram (2D) and JSmol (3D) viewer that are interlinked to show annotations in 1D, 2D and 3D.

Figure 4.

The ‘macromolecule’ view of the query interface. The results show the number of unique macromolecules (in this case 2262) present in the result set for the query ‘transferases’. A representative (or ‘best’) structure for each unique macromolecule is shown, with an image gallery highlighting the location of the particular macromolecule within the assembly. Other instances of this macromolecule in PDB entries are listed below. This is in addition to the traditional result interface which lists all the PDB entries that satisfy the given query criterion (≈18000 individual PDB entries in this case).