Matched Molecular Pair analysis – KNIME Community Hub

This workflow demonstrates how to to matched molecular pair analysis in KNIME and proposes a way how to deal with biological assay data that includes a qualifer. The data has been retrieved from ChEMBL27 and contains Ki values of hERG. The human ether-à-go-go-related gene (hERG) product is a potassium channel whose inhibition can cause potential lethal cardiac arrythmia, so introducing transformations to drug candidates that reduce hERG inhibition is desired. As MMP analysis results in a range of values, we will take the median to select transformations to investigate further. The Interactive View contains a table with all transformations that match the criteria defined by the user. When a transformation is selected, the details show up in a second table underneath.

This is a companion discussion topic for the original entry at

Thank you for the workflow!
It’s so useful and easy to understand.

I think, I found a little mistake in the “Configure the results of choice” component.

I typed the Number of occurrence = N (“Ioccurence-input”). And the MMPs with Number of occurrence = N were filtered out. So, I get MMPs only with (N+1). (For example, I needed MMPs with 5 numbers of pairs but get MMPs with 5+1=6 numbers.)


I found the reason in “Rule-Based Row Filter” node, the first node located in “Configure the results of choice” component:
Rule: $Number of occurences$ > $${Ioccurence-input}$$ => TRUE
Transformed rule: $Number of occurences$ >= $${Ioccurence-input}$$ => TRUE

But may be, it’s my misunderstanding :slight_smile:

Best regards,