Affinity Propagation Clustering

The affinity propagation algorithm takes as input a set of pairwise similarities among compound fingerprints, considering them as potential representative compounds (exemplars). The clusters and their corresponding representative compounds are calculated by exchanging messages between data points until a maximization process converge. Thus, exemplars for each cluster are proposed to the researcher for further investigation.

Step 1.
Please, Upload files* in ".sdf", ".mol", or binary fingerptint in ".txt" format.

*Warning:
‒ Uploaded filename should not contain special characters ^ ()_
‒ Files are temporary saved on the server and deleted after processing.
Maximum allowed upload size is 100MB .
fingerprint.txt format should contain two parts separated by a comma (,) character, the binary fingerprint and the name of the compound i.e:
00000110000001010000111101100000000000001111111111111110, Compound 0001
00111111111111111111000000000000000000110111000000000000, Compound 0002
01011110011001000110100100101011101110000000101111001101, Compound 0003
Binary part should not conatain any space before comma.
(see Example Data set 7. for more information)

Step 2. Please, Select Distance Parametres.

• Distance Selection:

Final Step.

Processing Data...
loading..
Please wait few minutes,
Do not Press Refresh or Any Other Button,
Results will be displayed automatically.


*After the next successful execution your downloadable file links will be:

1. SDF with exemplar compounds: http://chembioserver.vi-seem.eu/upload/TvGrrP4rZq_exemplars.sdf

*Results are stored for a week.