Homology Modeling

In homology modeling, the sequence is taken to BLAST and a specific template is searched for. Protein  BLAST is carried. The database option will be changed. The database by default is a non-redundant database but in this case, PDB is chosen so such sequences are picked up which are given in PDB.

BLAST can be joined with any database using its source code. Result of BLAST change as a result of changing the database.


Then a structure model can be predicted having sequence homology. The second name for homology modeling is Comparative modeling Of Protein as the model is comparative to the template. When the list of sequences comes then things like
  • Minimum e-value (around 0)
  • Maximum query coverage
  • Maximum percentage identity

help in determining the best template.

The sequence with 0 e-value can be chosen as the best homolog. When a sequence is BLAST searched through PDB then almost all entries have the same e-values around zero. The least characterized protein will have different e-values from zero. It also has to be seen how much the subject is being covered from the query and how much are identical positions.

The sequence with 100% percentage identity and 100% query coverage will have 0 e-value.

E-value is calculated by all these factors.

The subjects obtained against queries are called expected value. E-value can be positive, negative, or zero. It is calculated through value and being positive or negative depends on the denominator & numerator.

Homology modeling consists of four steps.

  1. Template Selection 
  2. Alignment
  3.  Model Modeling 
  4. Evaluating The Model

In the case of doing BLAST, steps 1 and 2 both occur simultaneously. The high chance of error is in the case of convergent evolution i.e

  • The similarity of sequence level but not structure.
  • The similarity of structure level but not function.

This is ignored by choosing more templates. Then multiple alignments are performed. Through tree, it is seen that even if sequence similarity is more then how much the two sequences are actually close. In multiple sequence alignment, it is assumed that out of all the templates with one is the closest relative. Thus, more than one models are mode. If the sequence has such a functional motif for a particular function then in which model it is correctly oriented.

 
Many other things like superimposition etc are carried to have the best match i.e which template is forming good angles. Sometimes those that are closely related are giving less homology and those that are distant are giving more.

In model building, some programs replace structure when homology comes e.g

A ➡️ Only Sequence.

B ➡️ Sequence + Structure.

The structure is replaced but amino acids are considered. The alpha-carbon trace or the peptide linkage chain is replaced. All proteins have the same alpha-carbon trace, differences come when side chains are added. The flow that structural dynamicity cannot be traced.

For example:
In the case of hemoglobin, valine replaces glutamic acid so the basic changed accrued in the side chain. Glutamic acid is negatively charged and forms some interaction due to which protein structure is in form. On the other hand, valine has an aliphatic side chain i-e inert/bulky. As a result form globule, the structure extends. The interaction of the heam group with peptide gets affected as a consequence of which oxygen-carrying capability is altered.

Thus, if modification occurs at a crucial point then structural change may occur. Therefore, while adding side chains when dynamicity is not seen then the structure will not match. In another approach, different side-chain considerations are taken into account. It is seen that which combination contains the minimum/lowest energy. When with different combinations, the energy range is close then different models are formed.

In the case of automated methods, only sequences are given, rest happens itself whereas in manual methods every step has to be taken.

There can be many functional states of the protein.
Combinations of modification form. This may be regarded as an environmental change but basically indirect genetic regulation. Change comes for the same time only.

Editor's Recommendation:

Homology Modeling Homology Modeling Reviewed by AB Ultimate Guides on June 09, 2020 Rating: 5

No comments:

Don't add any Spam link in comment box.

Powered by Blogger.