Evaluating Scoring Functions

A limitation of the original BLOSUM tests is that they did not assess the effect of gap scores. However, Pearson (1995) has comprehensively 

Johnson and Overington (1993) measured alignment accuracy and found good performance for structure-based, BLOSUM, and Gonnet matrices. As in other studies, log-odds matrices performed the best. 

The specific information present in a PSSM, even when it is made from only two or three homologous sequences, particularly favors the similarity between the sequences aligned to compute it and can improve searching sensitivity dramatically. This is because constraints on each position in a multiple alignment become better defined as more exam- 

---

An additional role of the model evaluation methods is to help in the actual modeling procedure. In principle, an improvement in the accuracy of a model is possible by incorporating the quality criteria into a scoring function being optimized to derive the model in the first place. 

Wang R, Lu Y, Wang S. Comparative evaluation of 11 scoring functions for molecular docking. J Med Chem 2003 46 2287-303. 

Now we turn to the scoring function. The function for evaluating the closeness of analogy involves a sensitivity parameter for the atom class, times the dissimilarity parameter, d ... 

Fig. 11.14. Schematic illustration of a COSMO-RS-based scoring function. Starting from the known surface polarities, a and a, of nearby receptor and substrate surfaces, the interaction energy could be evaluated using the COSMO-RS expressions for misfit and hb energies.

---