Major Databases

Nucleic (ENA, Genbank, DDBJ)

• ENA European Nucleotide Archive (go here to add DNA)
• they all share the same Accessionnumber

Protein (SwissProt, RefSeq, TREMBL)

• all proteins go through Interpro

Genomic (ENSEMBL)

• Older DB-Versions are always available
• only curated and publicized Genomes
• highly accurate
• Looking for a Species? go to ENSEMBL
• includes Cytogenic Banding (shows how stained Chromosomes have to look)
• EST shows were the gene is expressed (e. g. liver)

Protein structure (PDB)

DB for protein analysis

• Sequences from: Uniprot (Swissprot; trEMBL); Structure PDB (rcsb.org)
• Usually similar seq/structure = similar function
• Sequence motifs explained:

  {L} = everything except L; [LM] = M or L; x = everything [color=lightgreen]

• Use sequence profiles (e.g. motifs) to get more distant but related proteins, instead of just using BLAST
• Physiochemical properties (weight, electric load, hydrophili profile etc.) are taken into account for alignments

Secondary structure prediction

• use a combination of methods, not only one
• try get a consensus "prediction pattern" of all prediction methods

Best Site for this is "NPSA-pbil", its called consensus sec. structure prediction search here.

• other:

http://expasy.org/

http://www.unoriprot.g/

http://www.ebiokit.eu/

http://blast.ncbi.nlm.nih.gov/Blast.cgi

https://www.ebi.ac.uk/interpro

http://www.rcsb.org/pdb

• Protein Visualization with: UGENE Suite, DeepView, VMD (good one), Pymol, Rasmol/Jmol

Tridimensional structure prediction (homology modeling)

• You need to find a protein template (use BLAST or PSI-BLAST) with a similarity of at least 30 % and a structure
• Do a pairwaise alignment then
• Build a predicted structure with:
  • Rigid body assembly, segment matching (SWISS - MODEL, SegMOD)
  • Satisfaction of spatial restraints (distance and angles) => MODELLER, Geno3D
  • Artificial evolution (template - based methods with ab initio - like energy minimization principles) => NEST
• Check quality afterwards (PROCHECK, PDBSum etc. to see if the model seems possible)