Prediction-based fingerprints of protein–protein interactions
Interface definition
The concept of RSA plays an important role in the
subsequent definition of an interaction site and novel fingerprints of
protein interactions considered here.
RSA of the i-th amino acid residue, RSAi, is defined as the ratio of the solvent exposed surface area of that residue observed in a given structure, SAi, and some maximum value of the solvent exposed surface area for this kind of amino acid, MSAi
Hence, RSA adopts values between 0 and 100%, with 0% corresponding to a fully buried and 100% to a fully accessible residue.37
Unless specified otherwise (which may be relevant for comparison with RSA prediction methods from the literature), the maximum exposed surface areas are taken from,47 and correspond to those observed in an extended conformation of a tripeptide, with the residue of interest as the central residue. The DSSP program48 is used here to compute exposed surface areas. Also, unless specified otherwise, we define surface exposed residues as those that have RSA of 5% and more.
Specifically, an amino acid residue is regarded as an interaction site if
(i) it is surface-exposed when considering the structure of an individual protein chain;
(ii) the change in its RSA between the isolated chain and the corresponding complex structure is greater than 4% RSA, and the change in its exposed surface area in absolute terms is greater than 5 Å2.
As with other arbitrary thresholds, used in this work, we performed sensitivity analysis to assess the effects of such arbitrary choices and we also followed the literature as much as possible to enable comparison with other methods. In particular, the threshold of 4% RSA for the relative change in surface exposure between an isolated chain and complex structures corresponds to a typical error in RSA prediction for buried residues,35 which are most relevant for the novel fingerprints of protein interaction sites considered here. Moreover, this choice appears to be qualitatively consistent in terms of the resulting interaction interfaces with the work by Jones and Thornton7 and the resulting Protein–Protein Interaction Server for the analysis of protein complexes (http://www.biochem.ucl.ac.uk/bsm/PP/server/), as well as a more recent work by Offman and colleagues.52
RSA of the i-th amino acid residue, RSAi, is defined as the ratio of the solvent exposed surface area of that residue observed in a given structure, SAi, and some maximum value of the solvent exposed surface area for this kind of amino acid, MSAi
Unless specified otherwise (which may be relevant for comparison with RSA prediction methods from the literature), the maximum exposed surface areas are taken from,47 and correspond to those observed in an extended conformation of a tripeptide, with the residue of interest as the central residue. The DSSP program48 is used here to compute exposed surface areas. Also, unless specified otherwise, we define surface exposed residues as those that have RSA of 5% and more.
Definition of an Interaction Site
In this
work, following in the footsteps of previous studies we define
interaction sites based on the RSA change upon complex formation, that
is, RSA difference between an unbound and bound (complex) structure of
an individual chain.7, 51
For each chain considered here, its coordinates are first extracted from the corresponding complex structure and the DSSP program is used to compute the surface exposure of each amino acid residue in an unbound structure of a single protein chain.
Subsequently, residues that become buried at the interface upon complex formation may be identified by recomputing the level of surface exposure in the whole complex.
For each chain considered here, its coordinates are first extracted from the corresponding complex structure and the DSSP program is used to compute the surface exposure of each amino acid residue in an unbound structure of a single protein chain.
Subsequently, residues that become buried at the interface upon complex formation may be identified by recomputing the level of surface exposure in the whole complex.
Specifically, an amino acid residue is regarded as an interaction site if
(i) it is surface-exposed when considering the structure of an individual protein chain;
(ii) the change in its RSA between the isolated chain and the corresponding complex structure is greater than 4% RSA, and the change in its exposed surface area in absolute terms is greater than 5 Å2.
As with other arbitrary thresholds, used in this work, we performed sensitivity analysis to assess the effects of such arbitrary choices and we also followed the literature as much as possible to enable comparison with other methods. In particular, the threshold of 4% RSA for the relative change in surface exposure between an isolated chain and complex structures corresponds to a typical error in RSA prediction for buried residues,35 which are most relevant for the novel fingerprints of protein interaction sites considered here. Moreover, this choice appears to be qualitatively consistent in terms of the resulting interaction interfaces with the work by Jones and Thornton7 and the resulting Protein–Protein Interaction Server for the analysis of protein complexes (http://www.biochem.ucl.ac.uk/bsm/PP/server/), as well as a more recent work by Offman and colleagues.52
沒有留言:
張貼留言