UniProtKB/Swiss-Prot, the Manually Annotated Section of the UniProt KnowledgeBase: How to Use the Entry View.

By Emmanuel Boutet, Damien Lieberherr, Michael Tognolli, Michel Schneider, Parit Bansal, A. Bridge, S. Poux, L. Bougueleret, Ioannis Xenarios. 2016

795 citations recorded by the source corpus.

Subfields: Molecular Biology, Spectroscopy

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

In Spectroscopy

Citations in the corpus

The SWISS-MODEL Repository—new features and functionality (2016, background): “Information about the target protein includes the name of the target protein, links to UniProtKB (26), InterPro (27) and STRING (28).”
Modeling aspects of the language of life through transfer-learning protein sequences (2019, methodology): “The data set contained proteins from UniProtKB/Swiss-Prot [78] (release: 2016_04) with experimental annotation (code: ECO:0000269).”
Driven to Near-Experimental Accuracy by Refinement via Molecular Dynamics Simulations (2019, methodology): “Sequence profiles were generated with HHblits 29 by searching homologous sequences with the default options against Uniclust30, 30 which is a clustered UniProtKB database 31 at the level of 30% pairwise sequence identity, released in September 2016.”
DeepConv-DTI: Prediction of drug-target interactions via deep learning with convolution on protein sequences (2018, methodology): “For the proteins, we unified the identifiers of the proteins as UniProtKB/Swiss-Prot accessions [48].”
The language of proteins: NLP, machine learning & protein sequences (2021, background): “However, we lack reliable annotations for most proteins [17].”
CPC2: a fast and accurate coding potential calculator based on sequence intrinsic features (2017, methodology): “We selected protein-coding and non-coding transcripts that met rigorous criteria to obtain a testing set of high quality: for the protein-coding testing set, we obtained all nonpredicted mRNAs from the RefSeq database (24) with protein sequences annotated by Swiss-Prot (25) and redundant sequences (”
Protein language model embeddings for fast, accurate, alignment-free protein structure prediction (2021, background): “For instance OmpX from Escherichia Coli (Outer membrane protein protein X; Swiss-Prot identifier ompx_ecoli (Boutet et al., 2016)) has an 8-stranded beta-barrel.”
ProteinBERT: a universal deep-learning model of protein sequence and function (2021, methodology): “ProteinBERT was pretrained on ~106M proteins derived from UniProtKB/UniRef90, covering the entire tree of life [28, 29].”
Probabilistic variable-length segmentation of protein sequences for discriminative motif discovery (DiMotif) and sequence embedding (ProtVecX) (2018, methodology): “We trained the embedding on segmented sequences obtained from Monte Carlo sampling segmentation on the most recent version of the Swiss-Prot database [49], which contains 557,012 protein sequences.”
Modeling the language of life – Deep Learning Protein Sequences (2019, methodology): “During training, 38 each protein sequence in SwissProt [48] is split into overlapping 3-mers and another tool from 39 NLP, namely the word2vec (skip-gram) model [49] is used to predict adjacent words, given the 40 word at the center.”
Genome-wide prediction of disease variants with a deep protein language model (2022, methodology): “Protein language models are trained on a large set of protein sequences that have been selected throughout evolution (e.g. from UniProt [24]), thereby learning which sequence variations are more or less likely to occur.”
DeepGO: predicting protein functions from sequence and interactions using a deep ontology-aware classifier (2017, methodology): “We use SwissProt’s (Boutet et al., 2016) reviewed and manually annotated protein sequences with GO annotations downloaded on 05 January 2016 from http://www.uniprot.org/uniprot/.”
GOLabeler: Improving Sequence-based Large-scale Protein Function Prediction by Learning to Rank (2017, positive): “…labeled by GO terms of all ancestor nodes in the DAG; (ii) many labels per protein: we checked GO terms in Swissprot (Boutet et al. , 2016) (October 2016) with 66 841 proteins and found that a human protein is labeled by around 71 GO terms on average; (iii) large variation in the number of GO…”
COFACTOR: improved protein function prediction by combining structure, sequence and protein–protein interaction information (2017, background): “The confidence score for each predicted EC number is estimated based on the global and local similarity between the target and top template hit: CscoreEC = 2 1 + exp (− (0.25 × Lsim × SSas + TM + 2.5 × ID)) − 1 (8) where TM is the TM-score between query and template, ID is the sequence identity, SSa”