Other possible areas of study include the use of multiple similarity measures in the logistic regression model, accounting for individual judgements instead of using the majority decision, or the use of more sophisticated data fusion methods [22] than the simple consensus approach considered thus far.
From an engineering perspective, a hybrid model incorporating multiple similarity measures in combination, e.g. support vector machines (SVMs), is more popular than using individual models.
The application [ 98] of an ensemble method based on multiple similarity measures in combination with generalized boosted linear models to taxonomic marker (16S rRNA gene) profiles of the HMP cohort resulted in a global network of 3005 significant co-occurrence and co-exclusion relationships between 197 clades occurring throughout the human microbiome.
Thus the peripheral auditory and vestibular systems have multiple similarities, with some striking differences in the composition of accessory structures, support cells and hair cell fine structure [21].
Results In this work, the authors characterize ddTpp1, the Dictyostelium ortholog of TPP1, and show that ddTpp1 has multiple similarities to the human protein in proteolytic activity and trafficking.
In some studies, researchers have found that some specific types of headaches share multiple similarities [3-6].
Specifically, a molecule is more likely to be active the more frequently it is retrieved in multiple similarity searches using a single reference structure or in multiple similarity searches using structurally similar multiple reference structures.
Data fusion methods are widely used in virtual screening, and make the implicit assumption that the more often a molecule is retrieved in multiple similarity searches, the more likely it is to be active.
For the binding site independent TIFP, the Tanimoto similarity was not only calculated for the complex of the protein used for docking but individually for all reference complexes that were also used by PADIF scoring, resulting in multiple similarity based rankings for each dataset.
