MEDUSAMEDUSA (Multiclass flexibility prediction from sequences of amino acids) is a deep learning approach for prediction of protein flexibility from sequence. MEDUSA takes as input an amino acid sequence and returns a flexibility class of each residue in terms of the expected normalized B-factor value range. Prediction is simultaneously performed in two-, three- and five classes using a convolutional neural network trained on a dataset of non-redundant X-ray structures.https://www.dsimb.inserm.fr/MEDUSAmedusa-proteinbiotools:medusa-proteinWeb applicationhttp://edamontology.org/topic_2828X-ray diffractionhttp://edamontology.org/topic_0130Protein folding, stability and designhttp://edamontology.org/topic_0176Molecular dynamicshttp://edamontology.org/topic_0154Small moleculeshttp://edamontology.org/topic_0736Protein folds and structural domainsPythonPerlhttp://edamontology.org/operation_0244Simulation analysishttp://edamontology.org/operation_0474Protein structure predictionhttp://edamontology.org/operation_2476Molecular dynamicshttps://github.com/DSIMB/medusaRepositoryhttps://github.com/DSIMB/medusa/issuesIssue trackerhttps://hub.docker.com/r/dsimb/medusaContainer file10.1016/j.jmb.2021.16688233972018PrimaryTatiana Galochkinatatiana.galochkina@u-paris.frPersonPrimary contactJean-Christophe Gellyjean-christophe.gelly@u-paris.frPersonPrimary contact