CoChat
Sign Up
Cited by 107Protein Structure and DynamicsMachine Learning in BioinformaticsComputational Drug Discovery Methods

The DOI matches its Crossref record; 52 of the 53 references it cites were found in a registry or an index.

Prediction of hot spot residues at protein-protein interfaces by combining machine learning and energy-based methods

Open access
Lise, Stefano, Archambeau, Cedric, Pontil, Massimiliano, Jones, David T·2009·BMC Bioinformatics
DOI 10.1186/1471-2105-10-365PMC PMC2777894
Open this paper in CoChat — free accountOpen PDF

Abstract

BACKGROUND: Alanine scanning mutagenesis is a powerful experimental methodology for investigating the structural and energetic characteristics of protein complexes. Individual amino-acids are systematically mutated to alanine and changes in free energy of binding (DeltaDeltaG) measured. Several experiments have shown that protein-protein interactions are critically dependent on just a few residues ("hot spots") at the interface. Hot spots make a dominant contribution to the free energy of binding and if mutated they can disrupt the interaction. As mutagenesis studies require significant experimental efforts, there is a need for accurate and reliable computational methods. Such methods would also add to our understanding of the determinants of affinity and specificity in protein-protein recognition. RESULTS: We present a novel computational strategy to identify hot spot residues, given the structure of a complex. We consider the basic energetic terms that contribute to hot spot interactions, i.e. van der Waals potentials, solvation energy, hydrogen bonds and Coulomb electrostatics. We treat them as input features and use machine learning algorithms such as Support Vector Machines and Gaussian Processes to optimally combine and integrate them, based on a set of training examples of alanine mutations. We show that our approach is effective in predicting hot spots and it compares favourably to other available methods. In particular we find the best performances using Transductive Support Vector Machines, a semi-supervised learning scheme. When hot spots are defined as those residues for which DeltaDeltaG >or= 2 kcal/mol, our method achieves a precision and a recall respectively of 56% and 65%. CONCLUSION: We have developed an hybrid scheme in which energy terms are used as input features of machine learning models. This strategy combines the strengths of machine learning and energy-based methods. Although so far these two types of approaches have mainly been applied separately to biomolecular problems, the results of our investigation indicate that there are substantial benefits to be gained by their integration.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1186/1471-2105-10-365

Cites

52 of 52 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Schematic overview of protein structural regions which define the different energy contributions. The red filled area, (a), corresponds to side-chain atoms of the mutated residue; the red and blue striped regions, (b) and (c) respectively, correspond to atoms within 10 Å of the Cβ of the mutated residue. We distinguish 3 types of interactions: side-chain inter-molecular between (a) and (c), environment inter-molecular between (b) and (c), side-chain intra-molecular between (a) and (b).
Figure 1 Schematic overview of protein structural regions which define the different energy contributions. The red filled area, (a), corresponds to side-chain atoms of the mutated residue; the red and blue striped regions, (b) and (c) respectively, correspond to atoms within 10 Å of the Cβ of the mutated residue. We distinguish 3 types of interactions: side-chain inter-molecular between (a) and (c), environment inter-molecular between (b) and (c), side-chain intra-molecular between (a) and (b).CC BY · PubMed Central
Figure 2 Example of a lysine hot spot. A detail of the interaction between trypsin (in grey, surface representation) and trypsin inhibitor (in red) is shown (pdb code 2PTC). The side chain of the hot spot Lys15 from the inhibitor (in magenta, ΔΔG = 10 kcal/mol) fits into a hole on the trypsin surface. The ζ-nitrogen of the Lys side chain forms two hydrogen bonds (highlighted in green) with the residue Ser190 from trypsin.
Figure 2 Example of a lysine hot spot. A detail of the interaction between trypsin (in grey, surface representation) and trypsin inhibitor (in red) is shown (pdb code 2PTC). The side chain of the hot spot Lys15 from the inhibitor (in magenta, ΔΔG = 10 kcal/mol) fits into a hole on the trypsin surface. The ζ-nitrogen of the Lys side chain forms two hydrogen bonds (highlighted in green) with the residue Ser190 from trypsin.CC BY · PubMed Central
Figure 3 Examples of tyrosine hot spots. A detail of the interaction between colicin E9 immunity protein (in red) and colicin E9 DNase (in grey, surface representation) is shown (pdb code 1BXI). The side chain of the hot spots Tyr54 (ΔΔG = 4.83 kcal/mol) and Tyr55 (ΔΔG = 4.63 kcal/mol) are highlighted by a stick representation and are coloured in orange and magenta respectively. They fit into a groove of the E9 DNase surface. The residue Phe86 of E9 DNase is also displayed by sticks (dark grey with transparent surface representation). The hydroxyl group of Tyr54 is hydrogen bonded with the main-chain oxygen of Phe86 (green line); the aromatic ring of Tyr55 forms a stacking interaction with the aromatic ring of Phe86.
Figure 3 Examples of tyrosine hot spots. A detail of the interaction between colicin E9 immunity protein (in red) and colicin E9 DNase (in grey, surface representation) is shown (pdb code 1BXI). The side chain of the hot spots Tyr54 (ΔΔG = 4.83 kcal/mol) and Tyr55 (ΔΔG = 4.63 kcal/mol) are highlighted by a stick representation and are coloured in orange and magenta respectively. They fit into a groove of the E9 DNase surface. The residue Phe86 of E9 DNase is also displayed by sticks (dark grey with transparent surface representation). The hydroxyl group of Tyr54 is hydrogen bonded with the main-chain oxygen of Phe86 (green line); the aromatic ring of Tyr55 forms a stacking interaction with the aromatic ring of Phe86.CC BY · PubMed Central
Figure 4 Predicted versus observed changes in binding free energy ΔΔG. In blue are the predictions from SVM regression, in red those from Robetta. The continuous lines correspond to linear fits between ΔΔGcalc and ΔΔGobs, colour coded as the predictions (the respective equations are reported in the legend box).
Figure 4 Predicted versus observed changes in binding free energy ΔΔG. In blue are the predictions from SVM regression, in red those from Robetta. The continuous lines correspond to linear fits between ΔΔGcalc and ΔΔGobs, colour coded as the predictions (the respective equations are reported in the legend box).CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Cited by

Sorted by citations

See all 107

Related

See all

Metadata

Authors
Lise, S., Archambeau, C., Pontil, M., & Jones, D. T.
Affiliations
University College London (GB)
Corresponding author
Lise, Stefano
ORCID
Venue
BMC Bioinformatics
Publisher
Springer Science and Business Media LLC
Published
2009-10-30
Volume
10
Issue
1
Pages
365-365
ISSN
1471-2105 (ISSN-L)
DOI
10.1186/1471-2105-10-365
PMC
PMC2777894
OpenAlex
W2162402626
Cited by
107
References
53
Citations by year
2026: 3 · 2025: 1 · 2024: 4 · 2023: 1 · 2022: 8 · 2021: 5 · 2020: 5 · 2019: 2 · 2018: 10 · 2017: 5 · 2016: 7 · 2015: 7 · 2014: 8 · 2013: 12 · 2012: 8
FWCI
3.90
Type
journal article
Language
en
Open access
Licence
Locations
  • BMC Bioinformatics (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • UCL Discovery (University College London) (submittedVersion)
Funders
Biotechnology and Biological Sciences Research Council (BB/E017452/1)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
  • Life Sciences › Pharmacology, Toxicology and Pharmaceutics › Drug Discovery › Computational Drug Discovery Methods
Keywords
protein-protein interfaces, hot spot residues, alanine scanning, binding free energy, protein-protein interaction, energy-based features, van der Waals interactions, solvation energy, hydrogen bonding, electrostatics, support vector machine, Gaussian processes, semi-supervised learning
MeSH terms
Alanine, Artificial Intelligence, Binding Sites, Hydrogen Bonding, Proteins, Thermodynamics, Computational Biology, Protein Interaction Mapping, Databases, Protein, Static Electricity

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review