 Day students now will start reviewing what we have studied in structural prediction and homology modeling. So to begin with there are three strategies for predicting protein structures by using the sequences of the proteins. The sequences of the proteins were obviously obtained from either admin degradation or mass spectrometry. The three strategies that are available for you to predict the structure, the possible structure, that this amino acid sequence can take are homology modeling, full recognition or threading as it is called, and lastly ab initio methods. So we are going to review the homology modeling in this module and for the other two the other modules. So to start with you obtain the DNA sequence and you translated it into the amino acid primary sequence. If you remember the proteins had primary, secondary, tertiary and quaternary structures and you could obtain the sequence, the primary sequence using mass spectrometry. Next you search for sequence homologs and construct an alignment from the database. So this involves last or faster and you have a sequence db search. So once you have found the homologs that is the proteins which are similar in sequence then you go for finding the structure, the 3D structure of such homologs. The strategy is therefore called homology modeling. So once again you went for searching similar sequences which means with high alignment and identity and for those sequences you went to the PDB database and you found out the structures for each one of those sequences and therefore you drew the homology modeling pipeline. What if these 3D structures were not available then obviously we had another strategy that was the full assignment. So here let's discuss homology modeling the steps that are involved in this. So you may remember that there were seven very very important steps in homology modeling. So the first one was the template recognition and initial alignment. So this involved the identification of 3D structures by looking at their sequences and then you aligned your sequence with the sequence of the protein whose structure is unknown. Okay so once you had the initial alignment then you corrected your alignment by moving the gaps and the insertions and deletions. Following that you created the backbone that is the backbone of the protein and you inserted loops for regions that were missing like this and you then inserted the side chains onto each amino acid. So as you can see your protein is getting prepared really nicely and once the entire structure was complete you went for optimizing this model so you can use the MD simulations the molecular dynamic simulations and lastly the last step of course is model validation. So you had to validate the structure that it is biologically plausible. So in conclusion homology modeling had seven steps and it was one of the most important strategies for predicting structures of protein and the protein sequences. So it is only performed in cases where you have a high identity and alignment. That means that if you were to plot as you can see here in this small graph that homology modeling is useful if the identity is high and the alignment is high. So this is the homology modeling area but for region that is on the other side of the graph that is this portion which is also called the twilight zone you have some other strategies that we will discuss in the later modules.