Validation
Open3DALIGN was validated against two distinct validation suites based on the eight datasets gathered by Sutherland at al. [1].
First validation
- The original alignments were randomly scrambled
- Each of the three Open3DALIGN algorithms (namely, mixed, atom-based and pharmacophore-based) was independently applied on the scrambled datasets
- The performance of each algorithm was assessed in terms of RMSD from the original and of time required to carry out the superimposition
Second validation
- A 5-ns QMD search was carried out on each compound of the eight datasets
- The three Open3DALIGN methods were independently applied on the datasets, picking for each compound from the respective conformational database the conformer having the best match with the template. Compared to the first one, this is a much tougher validation run, since it challenges not only the ability of the algorithm to obtain a good match, but also the ability of the scoring function to actually identify the best matching conformer among many
- The performance of each algorithm was assessed in terms of RMSD from the original and of time required to carry out the superimposition
The whole validation suite can be downloaded here. To run it on your machine, please check the
README included in the tarball.
REFERENCES
- Sutherland, J. J.; O'Brien, L. A.; Weaver, D. F. J. Med. Chem. 2004, 47, 5541-5554 DOI
