Musite customized training fails when reading training files and training the model

[Nermin Pinar Karabulut]

Hi,

I want to use Musite for the benchmark analysis. I have a different kind of problem definition, thats why I want to use the customized prediction model. I created fasta files of phosphorylated proteins in my dataset, and used "#" after each occurrence of a phosphorylation site in sequences. I created all files in the same way, but for some of training files it sometimes fails when reading the file. Even it reads the file successfully, after disorder prediction it fails to train the model. Since there is not any explanation associated with the errors I mentioned, I cannot figure out the reasons. Do you have any suggestions about the possible reasons of these errors? Any help would be greatly appreciated.

Best,

Pinar