US EPA

2797619 

Journal Article 

A guest molecule-host cavity fitting algorithm to mine PDB for small molecule targets 

Byrem, WC; Armstead, SC; Kobayashi, S; Eckenhoff, RG; Eckmann, DM 

2006 

1 

Biochimica et Biophysica Acta
ISSN: 0006-3002
EISSN: 1878-2434 

1764 

8 

1320-1324 

English 

16904958 

10.1016/j.bbapap.2006.06.009 

WOS:000240537900002 

Inhaled anesthetic molecule occupancy of a protein internal cavity depends in part on the volumes of the guest molecule and the host site. Current algorithms to determine volume and surface area of cavities in proteins whose structures have been determined and cataloged make no allowance for shape or small degrees of shape adjustment to accommodate a guest. We developed an algorithm to determine spheroid dimensions matching cavity volume and surface area and applied it to screen the cavities of 6,658 nonredundant structures stored in the Protein Data Bank (PDB) for potential targets of halothane (2-bromo-2-chloro-1,1,1-trifluoroethane). Our algorithm determined sizes of prolate and oblate spheroids matching dimensions of each cavity found. If those spheroids could accommodate halothane (radius 2.91 A) as a guest, we determined the packing coefficient. 394,766 total cavities were identified. Of 58,681 cavities satisfying the fit criteria for halothane, 11,902 cavities had packing coefficients in the range of 0.46-0.64. This represents 20.3% of cavities large enough to hold halothane, 3.0% of all cavities processed, and found in 2,432 protein structures. Our algorithm incorporates shape dependence to screen guest-host relationships for potential small molecule occupancy of protein cavities. Proteins with large numbers of such cavities are more likely to be functionally altered by halothane. 

host cavity; protein cavity; spheroid; prolate; oblate; guest; guest molecule; algorithm; halothane; CASTp