Source code and data: (1) | Related publications: []
emb2dis is a deep learning model to identify disordered regions in protein sequences. This demo provides pre-trained models with ESM2 and ProtT5 embeddings on the DisProt database, ready to be used for disorder prediction on a single protein up to 1,000 amino acids. To process longer sequences you can install emb2dis from the repository.
Input sequences should be entered without ID. Sample input:
MSDNGPQNQRNAPRITFGGPSDSTGSNQNGERSGARSKQRRPQGLPNNTASWFTALTQHGKEDLKFPRGQGVPINTNSSP
DDQIGYYRRATRRIRGGDGKMKDLSPRWYFYYLGTGPEAGLPYGANKDGIIWVATEGALNTPKDHIGTRNPANNAAIVLQLPQ
GTTLPKGFYAEGSRGGSQASSRSSSRSRNSSRNSTPGSSRGTSPARMAGNGGDAALALLLLDRLNQLESKMSGKGQQQQGQ
TVTKKSAAEASKKPRQKRTATKAYNVTQAFGRRGPEQTQGNFGDQELIRQGTDYKHWPQIAQFAPSASAFFGMSRIGMEVTP
SGTWLTYTGAIKLDDKDPNFKDQVILLNKHIDAYKTFPPTEPKKDKKKKADETQALPQRQKKQQTVTLLPAADLDDFSKQLQQ
SMSSADSTQA
Contact: Sofia A. Duarte