![](img/p.png)
UniPPI: protein–protein interactions
UniPPI utilizes a neural network model based on the protein representation method UniRep .This model demonstrates excellent performance on protein-protein interaction data from yeast, Helicobacter pylori, and human proteins. Therefore, by collecting over 4 million protein-protein interaction data from the DIP, BIOGRID, and STRING databases, we have constructed a prediction model and provided a protein-protein interaction prediction platform, offering valuable insights for protein interaction studies.
![](img/model3.png)
Note:
1. The input should be protein sequences or in FASTA format.
2. The order of the protein sequences does not matter.
3. The prediction may take long depending on concurrent requests.