I am trying to validate tools for ab initio modelling for an M.Sc. thesis since my target proteins have no close homologues. I recently modelled three continuous fragments of a protein that has a crystal structure in PDB using QUARK (Job Q24734 and Job Q24784) still waiting for the final fragment.
Analysis by TM_Align for Job Q24734 superimposed on the crystal structure>>>>>>>>>>
-------------------------------------------------------------------------------------------------------------------
Name of Chain_1: A50579
Name of Chain_2: B50579
Length of Chain_1: 583 residues
Length of Chain_2: 200 residues
Aligned length= 134, RMSD= 4.58, Seq_ID=n_identical/n_aligned= 0.418
TM-score= 0.18820 (if normalized by length of Chain_1)
TM-score= 0.45541 (if normalized by length of Chain_2)
---------------------------------------------------------------------------------------------------------------------
1) Based on these two servers, FG-MD and MOdRefiner, can I use them to modify the model?
2)If the answer to question 1 is YES.......since this sequence has a crystal structure already, while my target protein doesn't, can I rely on the refinement process if I decide to model my target protein using Quark or what "tweaking" should I employ to get the best possible model?
3) Since Quark can only handle 200amino acids...is there i final linking process one can use to join two or more models from a continuous sequence?