Introduction
Homologous modeling, also known as comparative modeling of protein, is a useful routine to construct atomic-resolution model of the "target" protein from its amino acid sequence and an experimental 3D structure of a related homologous protein (the "template"). It is observed that protein tertiary structure is better conserved than amino acid sequence. Thus, even proteins that have diverged appreciably in sequence but still share detectable similarity will also share common structural properties. Moreover, it is difficult and time-consuming to obtain experimental structures from methods such as X-ray crystallography and protein NMR for every protein of interest, homology modeling can provide useful structural models for generating hypotheses about a protein's function and directing further experimental work.
In this exercise, we reproduce the ferredoxin model and structural quality analysis. We will then try to generate a better model, as assessed by ERRAT, by modifying the alignment. Two proteins, 1FXD and 1FDN, are used as templates.
Methods
Given the sequence of the templates, perform Basic Local Alignment Search Tool (BLAST) to find the most similar protein ferredoxin. Then, search ferredoxin through NCBI. As the templates 1FXD and 1FDN are 55-60 amino acids long, in order to obtain better results, choose the ferredoxin found in Streptococcus pyogenes since its length is about the same as the two templates. Click FASTA to obtain its sequence. Since we've got all the input needed, we can perform homologous modeling with the help of MODELLER, a computer program used in producing homology model of protein tertiary structures.
Open the terminal, create a file containing the target sequence using the command "vi target.ali". Paste the target sequence and convert FASTA format into PIR format. Note that an asterisk (*) should be included at the end of the target sequence. Save the file. After that, create a Python script to generate the alignment of the target sequence and two templates. In the terminal, type "vi alignment.py" to insert the script. Run the script using the command “mod9.14 alignment.py”. A file named "target-alignment.ali" will be generated. This file consists of three parts, with the two templates on top and the target sequence on bottom. Gaps are inserted to obtain the optimal alignment. The next step is to create a model.
Type the command "vi model.py" to insert the script. Create 5 models and assess those using DOPE score and GA341. The lower DOPE score, the better the model is. Specify chain A of two templates by adding letter "A" after each of them. After running the script, several files are generated, including a file named "model.log". Use the command "tail -15" to output the last 15 lines of the file "model.log". According to the DOPE score, model 5 is the best model and model 2 is the worst.
Results
Open VMD and load the 5 models and one of the templates 1FDN. Find the RMSD using RMSD Tools. Results show that model 3 has the least RMS deviations. Click Ramachandran Plot to see the secondary structure of model 3 and 1FDN. Both graphs show that the model 3 and 1FDN have alpha-beta folds, most amino acid are found on the beta strand. Repeat the same procedure with another template 1FXD. Model 3 is also the best model and it also shows alpha-beta folds and majority of amino acids are found on beat strand.
Alignment of 5 models and template 1FDN
Alignment of 5 models and template 1FXD
ERRAT, a protein structure validation program, is used to assess the backbone conformations and side-chain packing using statistical observations of the typical geometry of proteins. The program examines a PDB file, and generates a score based on the quality of the local structure surrounding each residue, as compared to the typical ranges of dihedral angles and side chain contacts observed in real proteins. Upload PDB file “target.B99990001.pdb” using the web form.
