Sentence examples for bond depth from inspiring English sources

2D topological fingerprints calculated to a bond depth of 4-6 contain sufficient information to allow the identification of SoMs using classifiers based on relatively small data sets.

Graphic to represent the construction of a circular fingerprint to a bond depth, d, of 6 using counts of SYBYL atom types at each level.

Classification performance on a top-2% basis shows similar trends although a plateau in classification performance tends to be reached earlier at a bond depth of 4-5.

Classification performance on an MCC basis is shown in Table 1 and generally improves with circular fingerprint bond depth since a more detailed description of each atomic site is being used, although the rate of improvement tends to slow with bond depth showing that the more local descriptors are most important.

Graphic to summarise the classification results for four molecules chosen at random from the CYP 3A4 data set using the scenario with a bond depth of 4, q=45, j=201 and the RASCAL classifier.

Figure 3 summarises the classification results for four molecules chosen at random from the CYP 3A4 data set using the scenario with a bond depth of 4, q=45, j=201 and the RASCAL classifier.

Graphs showing classification performance for the three data sets in terms of MCC against subsample length q for models built using circular fingerprint bond depths ranging from 4-6.

Graphs showing ROC curves for the three data sets using the best classification models built using circular fingerprint bond depths ranging from 4-6.

In this work we are concerned with classification problems where the elements x i l can assume integer values being the counts of SYBYL atom types at specific bond depths from the atom described by x i.

Graphs showing classification performance for the three data sets in terms of MCC against number of subsamples j for models built using circular fingerprint bond depths ranging from 4-6.

The SoM classification performance of each approach described in the Methods section was assessed for each isoform data set using feature vectors calculated at each of the bond depths ranging from 0-8.