Fingerprints

Hi,

To a table containing the content of an SDF file I've generated CDK/Fingerprints. When I try to use them as input for CDK/Fingerprint Simularity I get the error that a column of the value "BitVectorValue" is not available. The output from Fingerprints is of the newer type DenseBitVector.

Is there a way to use any Fingerprint similarity until this has been fixed?

 

Cheers,

Hendrik

Hi Hendrik,

both input ports accept values of type "DenseBitVectorCell" and should auto-configure as soon as you open the config window. If you use the CDK Fingerprinter KNIME node, it should work without throwing a configuration error. Can you please describe your problem in more detail or send me your workflow?

You can use also use Erl Wood Cheminformatics Fingerprint Similarity node (community repo).

Thanks,

Stephan

Hi Stephan,

Sorry for the late answer, was offline for a while. On my end, it seems that both Erl wood and CDK Fingerprint Similarity nodes require BitVectorValue as inputs - that's what the error messages say when I want to configure either node.

I've attached a workflow that uses both nodes, but since it errors in both cases you can't configure the last step. I've updated all nodes just before sending this comment; the whole KNIME installation was set up from scratch three weeks ago (KNIME 2.5.1, Win). Please let me know if you need further details.

 

Best,

Hendrik

Hi Hendrik,

no worries, thanks for the workflow.

Both fingerprint similarity nodes have two input ports and hence require two connections. You get an error message because the node cannot find a second BitVectorValue in the second (missing) connection.

For example, if you use a Row Splitter to split the output table of your Fingerprints node and following connect each output port to one input port of the Fingerprint Similarity node, the node should configure okay. I have attached the modified workflow.

Please note that the CDK KNIME fingerprint similarity integration node only supports MACCS fingerprints so far.

Hope that helps.

Stephan 

1 Like

Hi Stephan,

Thanks for this, now I understand it better.

Hendrik