Both are the valid methods to develop a tertiary structure of protein. The selection of the method depends upon the percentage identity of the query with that of the template. Check how much is the percentage match with the template and if its more matching its valid to take homology one. Also the selection of the final model depends on the validating results from Ramachandran Plot, ProSA analysis, Errat plot of the built model.

All these results should be taken into consideration and analysed thereby selecting a valid model out of the two methods followed.

Do not trust any of them much. The biggest issue till date with protein structure prediction servers is to predict all beta proteins. Plus what the homology modeling servers are going to do is that its going to use the Supra-35 structure from PDB and make yours look exactly like it. De novo servers are generally very bad at predicting amyloid structure or any beta sheets. Plus, all these protein structure prediction algorithms were designed to predict the structure of proteins in their equilibrium state and are not suitable for studying dynamics. Consider using simulations, but good luck with amyloid folding pathway, its a tough one.

Thank you all for your reply, I appreciate it a lot.

Dear Srayanta Mukherjee, do you know of any simulation software I could use to study Amyloid beta dynamics? I am new into this field, so I am in the process of getting to know the different tools available to progress with my research.

I suggest Maestro in the Schrodinger suite for molecular dynamics simulations. There are other software out there, but this is the one I used.

Hi Velmarini, may I suggest you read a old article that could help you to undertand Amyloid beta dynamics: Molecular modeling of the amyloid-beta-peptide using the homology to a fragment of triosephosphate isomerase that forms amyloid in vitro. Contreras CF, Canales MA, Alvarez A, De Ferrari GV, Inestrosa NC. Protein Eng. 1999 Nov;12(11):959-66.

Thanks the suggestions Mohannad and Carlos, I appreciate it.

Mohannad I've been told that GROMAC is another software I could use. I am new in this field , so which one you think would be easier for me to use first?

Not to sure why you would need to model Alzheimer's Amyloid Beta 1-42 fragment since there're several available crystal structures and NMRs available in the PDB. If you still want to continue modeling the structure then I would use several methods (you've mentioned a few already). Once you've collected a pool of 3D models (>100), you should use quality assessment methods (towards some consensus) to chose the top ~5. With some knowledge of the system you're interested in (in this case Amyloid Beta 1-42 fragment) you can determine whether these are close to valid structures. You can also use MD (like GROMACS) to see if there's a high RMSD from the initial pose.

I have never personally worked with GROMAC, but whichever gives more geometry options to monitor would be better. Plus, how I would do it is see which one is easier for me to operate as a program and go from there. All I can say is that maestro worked for me and it handled a homodimer for MD simulations, so I would take that into consideration.

You can use this experimental structure (PDB 2LFM) as a starting point for a simulation of an Abeta monomer (high resolution NMR structure of Abeta1-40 in buffer without detergents or organic solvents)

https://www.researchgate.net/publication/51464732_A_partially_folded_structure_of_amyloid-beta(1-40)_in_an_aqueous_environment?ev=prf_pub

However, it is very likely this NMR structure does not represent well the unordered conformations of Abeta, which do not give rise to strong NOEs.

Article A Partially Folded Structure of Amyloid-Beta (1-40) in an Aq...

If you plan to use MD or REMD with GB model, then there is an Abalone program.

http://www.biomolecular-modeling.com/Abalone/index.html

Hi, there are different structures in PDB in which beta-amyloid is fully/partial folded. I have used diverse crystal structures to understand the  mechanism in stabilization or disaggregation of formed fibrils starting to 1IYT and 2BEG, but there are other suitable models for computational studies. In my opinion is not necessary to use the homology modeling for this purpose. You can try to follow the events through MD simulation using a solved structure using DESMOND, GROMACS etc. For review you can see my articles

https://www.researchgate.net/publication/263165744_Disease-Modifying_Anti-Alzheimer%27s_Drugs_Inhibitors_of_Human_Cholinesterases_Interfering_with_-Amyloid_Aggregation

https://www.researchgate.net/publication/258149397_Multifunctional_Cholinesterase_and_Amyloid_Beta_Fibrillization_Modulators._Synthesis_and_Biological_Investigation

If you have some doubts and questions do not hesitate to contact me

Sincerely

Article Disease‐Modifying Anti‐Alzheimer's Drugs: Inhibitors of Huma...

Article Multifunctional Cholinesterase and Amyloid Beta Fibrillizati...