by Zsuzsanna Dosztanyi

Presentation

In PFD format: link

In PPT format: link

Tutorial

Exercise 1

DISPROT database and analyzing calcineurin A

  1. Find calcineurin A (PP2BA_HUMAN) in DISPROT www.disprot.org

    (DP00092)

    You can start from UNIPROT, it has a link to DISPROT or search DISPROT directly by keyword (sequence search does not work).

    1. What kind of information can you find on this page?
    2. Which regions are marked as disordered in DISPROT?
    3. Which regions are marked as ordered DISPROT?
    4. By what experimental techniques?
    5. What is the role of the disordered segments?

Exercise 2

Disorder prediction methods

  1. Collect prediction outputs for calcineurin A using various methods!

The input can be (depending on the method):

  • the amino acid sequence in FASTA format
  • amino acid sequence in raw format (without header)
  • UNIPROT ID or accession number

Please note

  • some methods are sensitive to line breaks.
  • In some cases the header cannot be too long or more then one word
  • There could be restrictions on minimum and maximum length of sequence

Some disorder prediction methods:

  1. Do the predictions agree with the experimental characterization of disorder?
  2. Do the predictions agree with each other?
  3. Which method predicts the most disorder?
  4. Note the differences in the running time of the methods.

Exercise 3

MobiDB database

  1. Find calcineurin A in MOBIDB http://mobidb.bio.unipd.it/entries/Q08209

You can find the output of several other methods there.

  1. Which regions are predicted as disordered by the majority of methods?
  2. Which regions have matching PDB structure?
  3. Can you find regions that cannot be unambigiously assigned? What could be the reason for that?

Exercise 4

Analyze the protein with the Uniprot acession: P05204

Find the corresponding Disprot entry (DP00039)

  1. Predict protein disorder for DISPROT DP00039, for example with IUPred
  2. Count number of positively charged amino acids
  3. Count number of negatively charged amino acids
  4. Calculate net charge (you can use the protparam server)
  5. Check low complexity segments
    • log on to the server machine using either putty (on WIndows) or your terninal (linux or Mac)
    • download the sequence using the following command:

wget http://www.uniprot.org/uniprot/P05204.fasta
This command downloads the sequence to your current directory - run seg command on your sequence, this will indicate low complexity regions in your sequence in lower case letters

seg P05204.fasta (or you can take these from PFAM through uniprot)

  1. Check PFAM domains

Is there a contradiction between PFAM domain assignments and predicted disorder?

Excercise 5

Identify likely disordered binding regions

Characterize viral protein (Q9IK92) from the viewpoint of disordered binding region

  1. Run ANCHOR for this protein (You can also run DISOPRED, but that will take longer)
  2. Can you find corresponding PDB structure for this protein? http://www.rcsb.org/pdb/protein/Q9IK92
  3. Compare predicted disordered binding regions with DISPROT annotation

Exercise 6

Filtering motif hits

Dynein light chain protein binds to disordered segments that have a TQT binding motifs. One of its known interaction partner is FA83D ( Q9H4H8 ) with the region VGTQTS. We found the same sequence fragment in the protein ASNSD1.

  1. Do you think it can be a valid binding partner?

Hint: Predict disordered binding regions (e.g. with ANCHOR)

You can add the VGTQTS motif to the search too in the motif window

Is the matching region predicted to have a disordered binding region?

Excercise 7

Characterize your favorite protein

in terms of disorder and disordered binding regions

If you don’t have a favourite protein, use N-WASP as an example (O00401)

  1. Run disorder prediction methods (e.g. IUPred, PONDR VSL2, ESpritz)
  2. Run disordered binding region prediction methods (e.g ANCHOR, DISOPRED3, MorfCHIBI)
  3. Collect Pfam annotation
  4. Check coiled coil and low complexity predictions
  5. Check MobiDB