Portal de Periódicos da CAPES

Tools 2021: New methods, new approaches, and a salute to recent developments in protein science

2020; Wiley; Volume: 30; Issue: 1 Linguagem: Inglês

10.1002/pro.4003

1469-896X

Brian W. Matthews,

Cancer Research and Treatments

It is now 3 years since we published our first “Tools” special issue. The outstanding success of that issue, in January 2018, prompted us to publish a follow-up issue 12 months ago. The present release, Tools 2021, is the third in the series and will now be followed on an annual basis. We especially thank the contributing authors for taking the time to share the tools they have developed with their colleagues in the field. During the time that I was a student at the University of Adelaide, Kendrew's group in Cambridge reported the first three-dimensional structure of any protein. Within my scientific lifetime that number has increased to over 170,000 structures in the Protein Data Bank. Notwithstanding the almost incomprehensible progress which has been made, one still needs to start with a physical sample of the protein. (Determining three-dimensional structure from an amino acid sequence still remains a challenge for the future.) Toward this necessary starting point, the paper from John Weir and coworkers (PRO-20-0146) describes an integrated bacterial and baculovirus expression vector suite. The system saves time by allowing simultaneous cloning into Escherichia coli and baculovirus expression vectors using the same PCR products. If a protein can be crystallized then X-ray crystallography is the obvious route to structure determination. The development of improved X-ray detectors, coupled with intense microbeams, has made it possible to use smaller and smaller crystals. Elspeth Garman and Joshua Dickerson (PRO-20-0100) discuss some of the issues associated with microcrystal techniques. How long will a crystal last before being ruined by the X-ray beam? How much damage is there not just from photoelectrons generated in the crystal but also from photoelectrons generated by material surrounding the crystal? The tool RADDOSE-3D helps the user choose the best conditions to optimize the quality of their X-ray data. There have been dramatic recent developments in electron microscopy, with structures recently determined at subatomic resolution. The group of Bridget Carragher and Clinton Potter have been leaders in developing techniques for automated collection of images from a transmission electron microscope. Here they describe new features and applications in their Leginon system (PRO-20-0183). J. Bernard Heymann developed Bsoft, a set of computational tools for processing electron tomographic tilt series to generate tomograms and interpret them. Here he updates recent implementations in Bsoft including segmentation-by-modeling (PRO-20-0139). There is an ongoing need for better methods to assess structure quality. Whenever a new protein structure is reported in the literature, the reader needs to ask “how reliable is this?”. When a structure is archived, the staff of the Protein Data Bank provide validation statistics. These are certainly helpful, especially for the experienced reader, but do not, for example, provide any advice as to whether the coordinates are reliable enough to justify molecular dynamics simulations, or ligand docking calculations. Here, William Rochira and Jon Agirre (PRO-20-0171) describe Iris, a tool that displays residue-by-residue quality, and highlights those parts of a protein model that may require attention. Bio3D is a family of packages with broad functionality including biomolecular database searching, sequence and structure analysis, and multivariate analysis methods, among others. Here Grant, Skaerven and Yao (PRO-20-0140) describe new methods for structure analysis and for dissecting sequence–structure–function relationships. One has to look no further than the cover of this special issue to appreciate that the visualization of protein structures is an art form. ChimeraX, developed by Thomas Goddard, Eric Pettersen, Thomas Ferrin and their coworkers (PRO-20-0131), is a widely used interactive visualization program. Such tools have become an indispensable part of the pursuit, visualization, communication and teaching of protein science. PyMOL is another very popular program for generating images of molecular structures. Here Blaine Mooers and Marina Brown (PRO-20-0231) describe a series of code templates which simplify and accelerate the use of PyMOL in generating images. ImageJ is a versatile image processing program that can visualize and process a wide range of biologically related images such as live-cell images, radiological images, and hematological images. Kevin Eliceiri and coworkers (PRO-20-0247) describe recent developments including plugins and tools to address user needs in several areas such as visualization, segmentation, and tracking of biological entities in large, complex data sets. Understanding and predicting protein stability and dynamics continues to be a key element of protein science. To facilitate measurement and analysis of protein stability, Marlovits and coworkers have developed MoltenProt and NanoDSF (PRO-20-0202). The former provides a robust way to characterize different types of protein samples including complexes and membrane proteins. The latter increases efficiency by using miniaturized measurement of intrinsic tryptophan fluorescence. DynaMut2 has been developed by David Ascher and coworkers (PRO-20-0133) to predict the effect of missense and other mutations on protein stability and dynamics. To avoid computationally expensive approaches the web server combines Normal Mode Analysis methods to capture protein motion with graph-based signatures for stability information. Doug Barrick's group describe a collection of programs for analysis of linear repeat proteins with point substitutions (PRO-20-0130). Rigorous statistical treatment is essential for Ising analysis of repeat proteins, especially heteropolymeric repeat proteins. Estimates of parameter uncertainty as well as correlations among fitted parameters are provided. Even in the earliest studies of protein structure it was noted that some surface side chains were disordered. No great significance was attached to this evidence for structural mobility. More recent studies, however, have shown that even within a single crystal, a protein may display alternative conformations that have functional significance. Here, van den Bedem and coworkers (PRO-20-0209) describe qFit 3, a procedure that can systematically identify and model alternate conformations seen in X-ray crystallographic and cryoelectron microscopic images of proteins. Even from the very first structure determinations it was apparent that proteins exist in families, and that knowledge of the structure of one family member (e.g., myoglobin) could be helpful in better understanding the structure of another family member (e.g., hemoglobin). The impact of that pioneering insight on the subsequent development of protein science would have been almost impossible to imagine. The first report in this section of the special issue relates to the SARS-CoV-2 family. Here, Wladek Minor's group have developed a web resource for SARS-CoV-2-related structural models (PRO-20-0174). There are already hundreds of experimental X-ray, cryo-EM, and NMR structures of proteins and nucleic acids related to this coronavirus. The website will allow novices and experts alike to navigate the flood of structural information. The MEROPS database was originally developed to systematize the classification and nomenclature of proteolytic enzymes. Here, Rawlings and Bateman (PRO-20-0132) discuss the extension of MEROPS to help understand peptidase specificity. The assembled substrate cleavage collection includes physiological, pathological, and nonphysiological cleavages in proteins, peptides, and synthetic substrates. Cecilia Lindskog and Andreas Digre describe recent additions to the Human Protein Atlas (PRO-20-0203). This is a large-scale initiative aimed at mapping the entire human proteome, specifying the exact location of proteins at tissue, cellular, or subcellular levels and its linkage to function. Use of antibody-based proteomics and various other omics technologies has provided a wealth of information supporting this exceptionally valuable database. Some protein complexes are recalcitrant to standard structure-determination techniques. Here Andrej Sali and coworkers describe an Integrative Modeling Platform (PRO-20-0233) to use multiple types of data to develop an ensemble of models consistent with the available information. Gunnar Jeschke (PRO-20-0137) also discusses MMM, his Integrative Modelling and Ensemble software. This includes approaches which can be brought to bear on systems where data come from different sources using different techniques. Sarel Fleishman and coauthors note that the functional sites of protein are often comprised of structured loop regions that are difficult to design, ab initio. They have therefore developed an approach called AbDesign (PRO-20-0134) that computationally assembles a large number of new backbones by combining naturally occurring modular fragments. This strategy can yield highly-functional proteins including antibodies and enzymes. The BioGRID resource, compiled by Mike Tyers and others (PRO-20-0181) is a manually curated database of protein and genetic interactions from multiple species. It also captures protein posttranslational modifications and protein or gene interactions with bioactive small molecules and drugs. For 30 years, AutoDock has provided tools for computational ligand docking and drug design. As described here by David Goodsell and collaborators (PRO-20-0129), it employs empirical force fields, docking engines, site prediction, and interactive tools for visualization and analysis. In order to better understand or utilize a particular metabolic pathway, it can be helpful to be able to introduce a “switch” which allows external control. The SynPHARM database, described by Jamie Davies (PRO-20-0224), is a software tool that helps engineer in a “normal” protein a drug-binding element whereby the activity of the protein can be externally controlled.