What aa should I use in 2nd structure prediction instead of X
1
1
Entering edit mode
15 months ago
Xylanaser ▴ 80

Im building pipe with secondary structure prediction (spider3). Program do not allow X in fasta sequence. What aa should I use in 2nd structure prediction instead of X? I think about G.

cheers

structure protein prediction • 515 views
3
Entering edit mode
15 months ago
Mensur Dlakic ★ 21k

What would be the rationale for using G? It has a clear propensity for random coil (more so than strand+helix combined), so there is a chance it will introduce something that is not there. My recommendation would be to use a residue that has almost equal propensity for all three states, and that will hopefully not distort the actual structure in that region. In order, those residues that can be found almost in any three states are: H, R and K. Below are relative propensities for all residues I mentioned.

A better suggestion may be to avoid using sequences with X residues, assuming you have that option. If not, a BLASTp search will tell you what best hits have at those X position, and then you replace them with actual residues from the best match.

    coil        strand      helix
G   1.701741723 0.675800457 0.49126143
H   1.12233694  1.0171953   0.896780032
K   0.980548113 0.807582702 1.170839916
R   0.892743354 0.926224811 1.192394115

0
Entering edit mode

thanks a lot!