About Dark Proteome Database  


Simplifying the generation of insight from protein structures

Nelson Perdigão1,2, Agostinho Rosa1,2, Seán I. O’Donoghue3

Updated: June 2016
This indicates when the Dark Proteome Database (DPD) was last calculated.
PDB structures released since then are not yet available in DPD.

1. University of Lisbon, Portugal
2. Instituto de Sistemas e Robótica, Lisboa, Portugal
3. Garvan Institute of Medical Research, Sydney, Australia

DPD Release 1.1.2016.06 - Current

DPD v1.1 covering
  • UniProt release 2016_06
  • PDB, Predict Protein, PMP, STRING 2016_06
  • Dark Proteome Map per Protein
  • Swiss-Prot sequences (550.116)
  • PDB structures (119.475)
  • Sequence to Structures ~ 49 M

DPD Release 1.0.2014.06

DPD v1.0 covering
  • UniProt release 2014_06
  • PDB, Predict Protein, PMP, STRING 2014_06
  • Swiss-Prot sequences (546.000)
  • PDB structures (100.326)
  • Sequence to Structures ~ 46 M

Curated Data

Annotations from the UniProt database include organism, subcellular location, tissue specificity, function, domains, and motifs;

Secondary structure is considered from the PDB whenever available;

Known interactors with evidence in databases and literature are taking into account.


PSSH2 homology database aligns 550.116 Swiss-Prot into 119.475 PDB generating ~ 49M 3D models;

We use Predict Protein to predict secondary structure, disordered and trans-membrane domains.

We use Protein Model Portal models to validate and complement PSSH2 sequence-structure alignments.

© Copyright 2015-2020 Nelson Perdigão and ISR-IST-ULisboa, All Rights Reserved.