The collection devices for Study 5 were not extracted using the "adhered cell" extraction developed for Study 2 due to the similarity of the observations for the initial extraction noted.
Since a complementary ridge was not extracted for every burst, we did not compute synchronization measures for this component.
The second is a split infrastructure security system, which lets key data such as passwords be entered in by an admin and used for login purposes, but not extracted back out by rank and file users.
For example, Bandicoot does not extract home location as a dummy variable in the feature space; however income is strongly correlated with home location (i.e. urban or rural).
In addition, 4 items that were extracted in the 2-group comparisons were not extracted for the 3-group comparisons (items 38, 49 and 51 out of the 22q11DS-ASD versus Heterogeneous ASD sample and item 62 out of the KS-ASD versus heterogeneous ASD comparisons).
Methods in square brackets yielded data that was not extracted for the purposes of this review.
We excluded studies for which we could not extract or reconstruct two by two tables, studies written in a language other than English, Dutch, German, or French, and reviews, editorials, and case reports.
Other relevant papers were not extracted but read for background information, and for placing some of the findings in a wider context.
An overview of examples for reasons why molecules are not extracted correctly can be found in the Additional file 4. In the patent→compounds use case only 63%% of the 2692 compounds found in SciFinder for the set of 46 patents were found in SureChEMBL, IBM SIIP or both.
