I. Determining the sequence number offset
Chains in PDB files are often numbered differently than the full-length chains in UniProt.
(MobiDB numbers the same as UniProt.) We will need to know the offset numbering difference.
I. Getting into MobiDB
- If there is more than one chain in the molecule you chose,
decide which chain you will analyze.
Generally, a longer chain,
or a chain with missing residues in the middle, will be more interesting.
In FirstGlance, you may wish to Isolate the chain. Then display it in Thin Backbone (Views tab).
Check Sequence Numbers and the "1" checkbox after Residue Names. Using
a white background, note the number
of the first (blue) residue in the chain. In the case of Chain A in 2arp, the first number is 1
(see snapshot at right).
Note the initial sequence, the first 6-8 residues.
For 2arp, the first 6 are GLECDG, and then there
are missing residues.
Make sure you get an initial sequence that is present in the genome-coded sequence.
Check your sequence alignment (Section 4). If some residues, such as a His tag, were
engineered onto the N-terminus, pick a sequence farther along that is identical with
the genome-encoded full-length sequence. Note the number of the first residue in
the sequence you use.
In FirstGlance, click on Sequences (Molecule Information Tab) and then get
UniProt sequence for your chain. Copy the sequence in FASTA format.
MSAReveal.Org. Paste your full-length FASTA sequence into the box, and click
the button Process Sequences below the box.
In the MSARevealed block, enter your leading sequence into the Find slot. The match in the
sequence listing will be highlighted in red.
- Touch the first red residue with your mouse and note the sequence number (see snapshot
at right). In this case, we see that
Gly 1 in chain A of 2arp is Gly 311 in the full-length sequence.
Thus, the offset is (311 - 1) = 310.*
When MobiDB tells you that disorder is predicted
for a sequence range, you will need to subtract 310 to get the corresponding sequence
range in 2arp.
Your offset will be different, and will be zero if you're lucky. Regardless, report
the offset in your slide like the *above statement.
II. What Each Bar Means
In FirstGlance in Jmol, click on Sequences (in the Molecule Information Tab), and
then on the link to the full-lenfth sequence in UniProt.
At UniProt, click on the blue Structure button at the left side of the page.
In the Structure section, look for MobiDb (near the bottom) and click on "Search...".
III. Predictions of Long Disordered Segments
- In the Overview section of MobiDB, here is how to interpret the sequence bars:
In the case above (3onz chain B) there are NO long segments of disorder predicted.
For short segments, only a few residues at each end are predicted to be disordered.
IV. Missing Residues vs. Predictions of Short Disordered Segments
If the bar for "Prediction of long disordered segments" has no orange regions
(as in the above snapshot), say in your slide
that no long disordered segments were predicted, and show a snapshot of that bar.
Then you can skip to section IV.
If any portions of this bar are orange, touch
them with your mouse and list the sequence ranges in your slide.
Below is an example that predicts long disordered segments.
Notice the black arrow on the first orange segment (representing a touch by the mouse).
A gray box pops up saying 50-96 Weak polyampholite (1). Intrinsicially disordered segments
are polyampholites, regions with a high density of both positively and negatively charged
- For the above example, the offset is zero. You would say in your slide:
Long disordered segments were predicted for 50-96, 282-325, and 351-393.
If the offset is not zero, you would add
In ["my homology model" or give the PDB code]
the predicted disordered segments are .... (numbers corrected with offset).
V. Disorder Predictions vs. Deleted Residues
- If you had orange segments predicting long disordered segments in the previous
section, you may SKIP what is below.
List the missing residues in the chain you are analyzing. You can copy this from
the report in FirstGlance.
If your sequence numbering offset is not zero, convert
the sequence numbers of missing residues to
those in the full-length sequence, and list those in your slide.
- In the bar "Prediction of disordered segements, short or long", touch the orange peaks
to see the sequence numbers. Are any of the missing residues predicted to be disordered with
a value >= 0.5? In the example at right, the black arrow (mouse) is touching sequence number 9.
A gray box appears saying "9 - 0.5" meaning that the disorder predition for residue 9 has a value
of 0.5. Some residues closer to the N-terminus have even higher values.
- In your slides, give the disorder prediction values for the missing residues.
Say which, if any, predictions agree with missing residues.
- Say in your slides whether residues that were deleted from the protein before it was
expressed, purified and crystallized are predicted to be disordered. (If you chose a homology
model, you can skip this.)
This page is maintained by