MetalDetector v2.0: predicting the geometry of metal binding sites from protein sequence

Passerini, Andrea; Lippi, Marco; Frasconi, Paolo

doi:10.1093/nar/gkr365

Cited by 62 publications

(57 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…3A), predicted to form disulfide bridges by Metaldetector (67). Outside of this region, we could detect no further sequence or secondary-structure similarity between ORF2 of CBPV and that of negeviruses.…”

Section: Resultsmentioning

confidence: 99%

Powerful Sequence Similarity Search Methods and In-Depth Manual Analyses Can Identify Remote Homologs in Many Apparently “Orphan” Viral Proteins

Kuchibhatla

Sherman

Chung

et al. 2014

J Virol

112

View full text Add to dashboard Cite

The genome sequences of new viruses often contain many “orphan” or “taxon-specific” proteins apparently lacking homologs. However, because viral proteins evolve very fast, commonly used sequence similarity detection methods such as BLAST may overlook homologs. We analyzed a data set of proteins from RNA viruses characterized as “genus specific” by BLAST. More powerful methods developed recently, such as HHblits or HHpred (available through web-based, user-friendly interfaces), could detect distant homologs of a quarter of these proteins, suggesting that these methods should be used to annotate viral genomes. In-depth manual analyses of a subset of the remaining sequences, guided by contextual information such as taxonomy, gene order, or domain cooccurrence, identified distant homologs of another third. Thus, a combination of powerful automated methods and manual analyses can uncover distant homologs of many proteins thought to be orphans. We expect these methodological results to be also applicable to cellular organisms, since they generally evolve much more slowly than RNA viruses. As an application, we reanalyzed the genome of a bee pathogen, Chronic bee paralysis virus (CBPV). We could identify homologs of most of its proteins thought to be orphans; in each case, identifying homologs provided functional clues. We discovered that CBPV encodes a domain homologous to the Alphavirus methyltransferase-guanylyltransferase; a putative membrane protein, SP24, with homologs in unrelated insect viruses and insect-transmitted plant viruses having different morphologies (cileviruses, higreviruses, blunerviruses, negeviruses); and a putative virion glycoprotein, ORF2, also found in negeviruses. SP24 and ORF2 are probably major structural components of the virions.

show abstract

Section: Resultsmentioning

confidence: 99%

Powerful Sequence Similarity Search Methods and In-Depth Manual Analyses Can Identify Remote Homologs in Many Apparently “Orphan” Viral Proteins

Kuchibhatla

Sherman

Chung

et al. 2014

J Virol

112

View full text Add to dashboard Cite

show abstract

“…These programs and services, including CMM, are listed in Supplementary Table 1, along with a comparison of the features of each. Programs that predict metal binding sites include FINDSITE-metal 35 , MetSite 36 , SVM-Prot 37 , SeqCHED 38 , and metalDetector 39 . Databases for querying metal binding sites include MESPEUS 16 , MIPS 40 , MDB 41 , MetalPDB 42 , Metal-MACiE 43 , MINAS (http://www.minas.uzh.ch/), and PROMISE 44 .…”

Section: Introductionmentioning

confidence: 99%

Validation of metal-binding sites in macromolecular structures with the CheckMyMetal web server

et al. 2013

View full text Add to dashboard Cite

Metals play vital roles in both the mechanism and architecture of biological macromolecules. Yet structures of metal-containing macromolecules where metals are misidentified and/or suboptimally modeled are abundant in the Protein Data Bank (PDB). This shows the need for a diagnostic tool to identify and correct such modeling problems with metal binding environments. The "CheckMyMetal" (CMM) web server (http://csgid.org/csgid/metal_sites/) is a sophisticated, user-friendly web-based method to evaluate metal binding sites in macromolecular structures in respect to 7350 metal binding sites observed in a benchmark dataset of 2304 high resolution crystal structures. The protocol outlines how the CMM server can be used to detect geometric and other irregularities in the structures of metal binding sites and alert researchers to potential errors in metal assignment. The protocol also gives practical guidelines for correcting problematic sites by modifying the metal binding environment and/or redefining metal identity in the PDB file. Several examples where this has led to meaningful results are described in the anticipated results section. CMM was designed for a broad audience—biomedical researchers studying metal-containing proteins and nucleic acids—but is equally well suited for structural biologists to validate new structures during modeling or refinement. The CMM server takes the coordinates of a metal-containing macromolecule structure in the PDB format as input and responds within a few seconds for a typical protein structure modeled with a few hundred amino acids.

show abstract

“…The MetalDetector software [5,6] uses state-of-the-art machine learning methods to solve the above three prediction problems. A first version of the software [5] employed Disulfind predictor to identify cysteine disulfide bridges [7] and a combination of support vector machines and bidirectional recurrent neural networks for metal bonding state prediction.…”

Section: Metal Binding In Proteinsmentioning

confidence: 99%

“…A first version of the software [5] employed Disulfind predictor to identify cysteine disulfide bridges [7] and a combination of support vector machines and bidirectional recurrent neural networks for metal bonding state prediction. The current version 2 of the server [6] employs a two-stage approach for metal bonding state and metal binding sites prediction respectively. The first stage relies on an SVM-HMM [8] which collectively assigns the bonding state of all the CYS/HIS residues in the sequence.…”

Section: Metal Binding In Proteinsmentioning

confidence: 99%

Metal Binding in Proteins: Machine Learning Complements X-Ray Absorption Spectroscopy

Lippi

Passerini

Punta

2012

Machine Learning and Knowledge Discovery in Databases

Self Cite

View full text Add to dashboard Cite

Abstract. We present an application of machine learning algorithms for the identification of metalloproteins and metal binding sites on a genome scale. An extensive evaluation conducted in combination with Xray absorption spectroscopy shows the great potentiality of the approach. Metal Binding in ProteinsA significant fraction of known proteins is believed to bind metal ions in their native conformation. Metal ions play a variety of crucial roles in proteins [1], from stabilizing their three dimensional structure, to acting as cofactors in enzyme catalysis. Moreover, metals are implicated in many diseases for which medicine is still seeking an effective treatment, such as Parkinson's or Alzheimer's [2] . Identifying unknown metalloproteins and detecting their metal binding site(s) is an important step in understanding their function and characterizing many crucial processes involved in living systems. A metal binding site is characterized through its metal ion, the protein amino acid residues directly involved in binding it (called ligands) and the binding geometry, i.e. the spatial arrangement of the ion and its ligands. Some metal binding sites actually involve compounds (e.g. the heme group binding hemoglobin) including one or more ions, and some proteins contain more than one metal binding site. The problem of identifying and characterizing metalloproteins can be seen as a series of increasingly complex tasks. Given a protein, one aims at: 1) determining whether it is a metalloprotein or not, i.e. if it binds metal ions in its native conformation; 2) determining the metal bonding state of each of its residues, i.e. whether they bind a metal ion or not; 3) determining the composition of metal binding sites, i.e. the number of ions binding the protein and the set of their respective ligands. Answers may be obtained experimentally, by means of in-silico prediction tools, or by a combination of the two classes of methods.Experimental Methods. High-throughput techniques based on X-ray absorption spectroscopy [3] (HT-XAS) allow detection, identification and quantification of metals bound to proteins based on the energy and intensity of the X-ray fluorescence signal emitted by the metals. HT-XAS can thus be used to address the first task, i.e. metalloprotein identification. Exact determination of binding sites, however, is only possible using more labor-intensive techniques, such as X-ray

show abstract

MetalDetector v2.0: predicting the geometry of metal binding sites from protein sequence

Cited by 62 publications

References 23 publications

Powerful Sequence Similarity Search Methods and In-Depth Manual Analyses Can Identify Remote Homologs in Many Apparently “Orphan” Viral Proteins

Powerful Sequence Similarity Search Methods and In-Depth Manual Analyses Can Identify Remote Homologs in Many Apparently “Orphan” Viral Proteins

Validation of metal-binding sites in macromolecular structures with the CheckMyMetal web server

Metal Binding in Proteins: Machine Learning Complements X-Ray Absorption Spectroscopy

Contact Info

Product

Resources

About