I would like to perform redundancy reduction for my test set for membrane topology prediction. I checked the 1998 paper and website of the HMMTOP server but haven't found any download links for the dataset that they have used for training.


In the Appendix Section of the paper, there is a table of entries and topologies. Below, there is a note:

*Entries which are included in the initial parameter and pseudocount settings from the 83TMP set.

Please correct me if I'm wrong, I suppose these those 63 protein entry names with an asterisk are the data used for training?

  • $\begingroup$ From materials and methods: The first data set was originally collected by Jones et al. (1994), and was also used by Rost et al. (1996) (83TMP set). The second one is an extension of it by Rost et al. (1996) (48TMP set). The third data set contains prokariotic transmembrane proteins collected by Cserző et al. (1997) (prokTMP set). $\endgroup$ – WYSIWYG Apr 25 '19 at 15:33
  • $\begingroup$ You might have to look at these papers individually or mail and ask the authors for a properly formatted dataset that they used for training. The last update was 18 years ago; so I'm not sure if the author is still active (alive)!! Perhaps you should try more recent tools based on HMMTOP. Check for citations. $\endgroup$ – WYSIWYG Apr 25 '19 at 15:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.