Hi, I'm beginner of protein modeling, and i-tasser is my first modeling program. So I used i-tasser suite on my computer(ubuntu 17.04) with some protein sequences, and I got problem on model selection problem.
Below is content of my cscore file.
Threading Zmax Znorm
MUSTER 4.10 0.71
dPPAS 5.26 0.57
wdPPAS 3.99 0.43
wMUSTER 3.37 0.58
wPPAS 3.13 0.45
dPPAS2 6.24 0.59
PPAS 3.72 0.53
Env-PPAS 6.16 0.77
==================================================================
| Estimated accuracy of your models |
==================================================================
Estimated Clusters (N=11295)
--------------------- ------------------
Model# C-score TM-score RMSD (A) #decoys density
------ ------- ---------- -------- ------- -------
model1 -4.46 0.25+-0.07 18.3+-2.3 753 0.020
model2 -3.73 753 0.042
model3 -5.00 473 0.005
model4 -5.00 466 0.005
model5 -5.00 462 0.005
First I saw this, I just think model 1 will be right. But when I saw the shape of it, it's shape was not reasonable compared to phylogenetic tree of my sequences. Model 2 was more reasonable.
So I think model 2 is right. But I cannot be sure about that. C-score, # of decoys, and even density of model 2 is better.
Can I select model 2 as right thing? Why model 1 was made as first rank?
P.S)I also wonder how TM-score and RMSD are calculated. I cannot found reference model in my result directory.