|
KNIME extension |
|
KNIME (http://www.knime.org) is an open-source data integration, processing, analysis, and exploration platform. DRAGON 6.0 Extension for KNIME is a package that allows to easily use the capabilities of Dragon 6.0 inside the KNIME environment.
Requirements The extension works as a wrapper for Dragon, thus it is needed a registered version of Dragon 6.0 (the extension does not work for previous version of Dragon). Please note that the extension uses the Command Line Interface of Dragon, that has to be registered separately from the Graphic User Interface version (refer to Dragon 6.0 manual for further details). The extension has been developed for Dragon version 6.0.13 or higher. If you have a previous version of Dragon you are recommended to upgrade it. The extension has been designed and developed using KNIME version 2.3.4. If you use a different version of KNIME and you experience problems, please contact us.
How to obtain the extension The extension is freely available for download on Talete website at http://www.talete.mi.it/products/knime.htm
How to install the extension To install the extension, put the file org.knime.dragon6_1.0.4.jar in the folder "dropins" located inside your main KNIME installation folder. If you have some older versions of the node file in the same directory, remove them. The extension will be available inside the KNIME environment at its next run.
How to use the extension The extension consists in a package of two nodes: Dragon 6.0 and File Reader. Please refer to the Node Description panel available in the KNIME environment for details on options, input and output ports and views of these nodes.
The Dragon 6.0 node takes a molecule file as input and produces a table with calculated descriptors as output. The input of the node has to be a table, in which the molecule file is contained in a column, then in the Dragon 6.0 node option you have to select the column containing the source of the molecule file. For example, if you have installed the KNIME Chemistry extension, the SDF Reader node can read a single SDF file and load it in a table, where the MOL source of each molecule is contained in each row of the column labeled "Molecule". If, for instance, you are willing to put Dragon after a molecule pre-processing work-flow, the node CDK to Molecule can be used, and it will produce the same output as the SDF Reader node. The provided File Reader node can also be used, it loads a generic molecule file into a table, putting each text row of the source in a single row.
After having set the input for the Dragon 6.0 node, two other node option should be set for its proper working:
All the other options of the Dragon 6.0 node let you arrange the calculation option as you would do in Dragon GUI: you can choose descriptors blocks to be calculated, choose the weighting schemes, and choose the output filters for descriptors and molecules. You can refer to the Dragon manual for details on the meaning of each option.
After the calculation has been performed, the Dragon 6.0 node provides as output a table containing the calculated descriptors (as columns) for each processed molecule (as rows). Furthermore, the log of the calculation is provided in the "Calculation log" view.
If some fatal errors were encountered, in the "Calculation log" view will be reported some information about the errors and the output of Dragon. Note that when such errors are encountered, the node will send as output an empty table; this will be clearly noticed as a warning icon should appear on the Dragon node, and the message “Node created an empty data table” will appear in the KNIME Console panel. Please consider that common fatal errors have the following causes:
|