different pocket results from same structure

[David Oppenheimer]

I have 2 pdb files that I have used for fpocket analysis. These are not identical files (the b-factor values are different), but when I check the RMSD in ChimeraX, the result is 0.000. However, when screened on fpocket web, file #1 shows the best pocket to be:
Pocket 1 :
Score : 0.459
Druggability Score : 0.798
Number of Alpha Spheres : 91

Whereas when file #2 is screened, it gives the best pocket as:
Pocket 1 :
Score : 0.665
Druggability Score : 0.503
Number of Alpha Spheres : 52

The pockets are in the same area of the protein, though.

These values seem significantly different from each other considering the RMSD of the 2 structures is 0. Do entries in the B-factor column of the pdb file affect the scoring?

Any insight welcome. Thank you for your help.

[Jacob Durrant]

Hi David. Thanks for reaching out re. your FPocketWeb results. Nothing in our implementation of fpocket (FPocketWeb) introduces a random element, so I would guess that fpocket itself, the program on which our project is based, is non-deterministic.

To get a more definitive answer, I recommend contacting the developers of the original fpocket program. They would have a deeper understanding of the code base and could provide insights into any non-deterministic factors that may influence the scoring differences you saw.

I hope this information helps! Thanks again for your interest in FPocketWeb.

[Author]

Posts