The following node is available in the Open Source KNIME predictive analytics and data mining platform version 2.7.1. Discover over 1000 other nodes, as well as enterprise functionality at http://knime.com.
This node draws enrichment curves often used in virtual screening. For this the user can choose a column by
which the data is sorted and represents the x-axis. The values on the y-axis are formed by the sum of the
hits in a second column that is also selected by the user. A row is considered a hit if the value is greater
than 0. The steeper the resulting curve, the better the enrichment is. Optionally the y-axis can show the
sum of the hit values instead of the number of hits.
The two gray lines in the view show the enrichment if the data points were ordered randomly (lower diagonal)
and the optimal enrichment if all hits are ordered before the first non-hit (upper diagonal).
0 | Input data with predicted values and actual values |
0 | A one-column table with the area(s) under the enrichment curve(s) |
1 | Table with the discovery rates (either hits or clusters) for the different curves at selected points (given as fraction of the complete dataset). |