In this article, we proposed a model named XG-PseU that can be used to identify pseudouridine sites. By analyzing the existing dataset and the latest database RMBase v2.0, we identified some differences and modified the original dataset. Compared our classifier with other's on 10-validation and independent datasets, the results showed that our model is more stable.
Our classifier can be used to identify pseudouridine site in H. sapiens, M. musculus and S. cerevisiae. For H. sapiens and M. musculus, please make sure the input RNA sequence contains no less than 21 nucleotides and in FASTA format; and S. cerevisiae the query sequence contains no less than 31 nucleotides with the FASTA format.