Filtering candidates by calculated properties

In this subsection, we look at how the calculated properties introduced in the sections above can be used to filter or eliminate structure candidates. This strategy was explored in [287] and is summarized below. First the general strategy is presented, followed by a theoretical example using several compounds of the formula C12H10O2. 

The number of calculated properties described in the previous sections require a logical approach for efficient and reproducible structure elucidation incorporating so many criteria. One approach is to use systematic filtering. This is best spUt into two basic parts structure generation (i.e. how to constrain the number of candidates using substructures and when to add additional constraints) and structure elimination (i.e. which is the best order to eliminate candidates using calculated properties). 

the use of the exclusion strategy is demonstrated using a theoretical example of several compounds with the same molecular formula. The formula C12H10O2 was chosen as there were several EI-MS spectra in the NIST [225] database that covered compounds with a wide range of properties, as well as isomers differing in substituent positions, which are very difficult to distinguish using MS information alone. In total. 

Before moving on any further, we first consider the effectiveness of the mass spectral classifiers in reducing the number of candidates for selection. Of the 29 compounds, 15 were already reduced to the final, small ( 35) number of candidates using 95% probability classifiers alone. No additional classifiers were necessary (or possible) to reduce this number further, nor did calculated properties separate these candidates further. This means over half these compounds required only the default substructure information to constrain structure generation adequately. In a further 4 cases, the reduction in candidates to their final number (again below 35) was achieved by using an additional classifier or other restrictions, and no further elimination occurred with calculated properties. One further compound experienced only a slight additional re- 

Although we gained many orders of magnitude in candidate reduction overall, some weaknesses remained with this exclusion strategy. Of greatest concern are inaccuracies, both in the classifiers and the calculated properties. As is evident from 

---