Please paste up to 5000 FASTA-formatted sequences in the textarea below:

Select mode: Original model   New model (using dipeptides)

Prior information (constrained prediction):

Note: If you wish to perform a batch submission of more than 5000 sequences, please contact directly.

- If you find HMM-TM useful in your research, please consider citing the reference that describes this work:
Algorithms for incorporating prior topological information in HMMs: application to transmembrane proteins
Pantelis G. Bagos, Theodore D. Liakopoulos and Stavros J. Hamodrakas
PMID: 16597327

- If you further use the new ("dipeptides") model, please also cite:
Extending Hidden Markov Models to allow conditioning on previous observations
Ioannis A. Tamposis, Margarita C. Theodoropoulou, Konstantinos D. Tsirigos and Pantelis G. Bagos
under review in Bioinformatics