DeepLoc

DeepLoc is a bioinformatics software tool that predicts the subcellular localization of eukaryotic proteins using deep neural networks. The software relies solely on sequence information and uses a recurrent neural network and an attention mechanism to identify important protein regions. DeepLoc was trained and tested on a protein dataset extracted from UniProt and achieved high accuracy (78% for 10 categories; 92% for membrane-bound or soluble), outperforming state-of-the-art algorithms that rely on homology information. DeepLoc is available as a web server, and example code and dataset are also available.

Topic

Sequence analysis;Protein properties

Detail

  • Operation: Protein subcellular localisation prediction

  • Software interface: Web service

  • Language: -

  • License: -

  • Cost: Free

  • Version name: 2.0

  • Credit: Novo Nordisk Foundation and the NVIDIA Corporation

  • Input: -

  • Output: Predicted localizations and signals, Probability table, Sorting signal importance

  • Contact: Henrik Nielsen hnielsen@cbs.dtu.dk,José Juan Almagro Armenteros jjalma@dtu.dk

  • Collection: -

  • Maturity: Stable

Publications

  • DeepLoc: prediction of protein subcellular localization using deep learning.
  • Almagro Armenteros JJ, et al. DeepLoc: prediction of protein subcellular localization using deep learning. DeepLoc: prediction of protein subcellular localization using deep learning. 2017; 33:3387-3395. doi: 10.1093/bioinformatics/btx431
  • https://doi.org/10.1093/bioinformatics/btx431
  • PMID: 29036616
  • PMC: -

Download and documentation


< Back to DB search