Limits...
Local functional descriptors for surface comparison based binding prediction.

Cipriano GM, Phillips GN, Gleicher M - BMC Bioinformatics (2012)

Bottom Line: Our approach uses a variety of physical, chemical, and geometric properties, adaptively weighting these properties as appropriate for different regions of the interface.Our approach builds a classifier based on a training corpus of examples of binding sites of the target ligand.The constructed classifiers can be applied to a query protein providing a probability for each position on the protein that the position is part of a binding interface.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Computer Sciences, University of Wisconsin-Madison, Madison, WI 53706, USA.

ABSTRACT

Background: Molecular recognition in proteins occurs due to appropriate arrangements of physical, chemical, and geometric properties of an atomic surface. Similar surface regions should create similar binding interfaces. Effective methods for comparing surface regions can be used in identifying similar regions, and to predict interactions without regard to the underlying structural scaffold that creates the surface.

Results: We present a new descriptor for protein functional surfaces and algorithms for using these descriptors to compare protein surface regions to identify ligand binding interfaces. Our approach uses descriptors of local regions of the surface, and assembles collections of matches to compare larger regions. Our approach uses a variety of physical, chemical, and geometric properties, adaptively weighting these properties as appropriate for different regions of the interface. Our approach builds a classifier based on a training corpus of examples of binding sites of the target ligand. The constructed classifiers can be applied to a query protein providing a probability for each position on the protein that the position is part of a binding interface. We demonstrate the effectiveness of the approach on a number of benchmarks, demonstrating performance that is comparable to the state-of-the-art, with an approach with more generality than these prior methods.

Conclusions: Local functional descriptors offer a new method for protein surface comparison that is sufficiently flexible to serve in a variety of applications.

Show MeSH
This image shows the charge (indicated by color, range from dark red (very negative) to dark blue (very positive) of protein 1AYP as computed by APBS. This charge pattern is quite different than any of the others seen in the training set.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3585919&req=5

Figure 10: This image shows the charge (indicated by color, range from dark red (very negative) to dark blue (very positive) of protein 1AYP as computed by APBS. This charge pattern is quite different than any of the others seen in the training set.

Mentions: The results of the calcium ion binding experiment are generally successful. In all but one test, our approach is able to localize the binding site well. Interestingly, as many of these sites are on the “outside” of the protein, it is unclear whether pocket-based methods would be as successful. The one failure case in the test (1AYP, shown in the right of Figure 1 and the lowest line in Figure 8a) is a protein unlike any of the 100 others in the training set as it has a positive electrostatic charge over most of its surface (Figure 10). This failure underscores the importance of finding a training set that adequately covers the potential interfaces. Our sampling procedure was designed to find such sets, but relevant examples may have been removed when we randomly reduced the training set size. Also, the sampling process relies on the availability of examples in the PDB and on clustering obtained by BLASTCLUST that considers sequence homology, not functional surface diversity.


Local functional descriptors for surface comparison based binding prediction.

Cipriano GM, Phillips GN, Gleicher M - BMC Bioinformatics (2012)

This image shows the charge (indicated by color, range from dark red (very negative) to dark blue (very positive) of protein 1AYP as computed by APBS. This charge pattern is quite different than any of the others seen in the training set.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3585919&req=5

Figure 10: This image shows the charge (indicated by color, range from dark red (very negative) to dark blue (very positive) of protein 1AYP as computed by APBS. This charge pattern is quite different than any of the others seen in the training set.
Mentions: The results of the calcium ion binding experiment are generally successful. In all but one test, our approach is able to localize the binding site well. Interestingly, as many of these sites are on the “outside” of the protein, it is unclear whether pocket-based methods would be as successful. The one failure case in the test (1AYP, shown in the right of Figure 1 and the lowest line in Figure 8a) is a protein unlike any of the 100 others in the training set as it has a positive electrostatic charge over most of its surface (Figure 10). This failure underscores the importance of finding a training set that adequately covers the potential interfaces. Our sampling procedure was designed to find such sets, but relevant examples may have been removed when we randomly reduced the training set size. Also, the sampling process relies on the availability of examples in the PDB and on clustering obtained by BLASTCLUST that considers sequence homology, not functional surface diversity.

Bottom Line: Our approach uses a variety of physical, chemical, and geometric properties, adaptively weighting these properties as appropriate for different regions of the interface.Our approach builds a classifier based on a training corpus of examples of binding sites of the target ligand.The constructed classifiers can be applied to a query protein providing a probability for each position on the protein that the position is part of a binding interface.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Computer Sciences, University of Wisconsin-Madison, Madison, WI 53706, USA.

ABSTRACT

Background: Molecular recognition in proteins occurs due to appropriate arrangements of physical, chemical, and geometric properties of an atomic surface. Similar surface regions should create similar binding interfaces. Effective methods for comparing surface regions can be used in identifying similar regions, and to predict interactions without regard to the underlying structural scaffold that creates the surface.

Results: We present a new descriptor for protein functional surfaces and algorithms for using these descriptors to compare protein surface regions to identify ligand binding interfaces. Our approach uses descriptors of local regions of the surface, and assembles collections of matches to compare larger regions. Our approach uses a variety of physical, chemical, and geometric properties, adaptively weighting these properties as appropriate for different regions of the interface. Our approach builds a classifier based on a training corpus of examples of binding sites of the target ligand. The constructed classifiers can be applied to a query protein providing a probability for each position on the protein that the position is part of a binding interface. We demonstrate the effectiveness of the approach on a number of benchmarks, demonstrating performance that is comparable to the state-of-the-art, with an approach with more generality than these prior methods.

Conclusions: Local functional descriptors offer a new method for protein surface comparison that is sufficiently flexible to serve in a variety of applications.

Show MeSH