 Dear students, now we are going to discuss the advantages and disadvantages of threading. Threading or fold recognition as it is also known as is a process by which you can predict the tertiary structures of proteins by looking at the sequence of a protein and comparing it with a template. Now if you were given a protein sequence and you wanted to predict its structure by using threading then all you had to do was to compare the secondary structures that could be made from that sequence with the fold database or the secondary structure database and thereby you would arrive at the possibilities or the best matches for the secondary structures which could be threaded on to the sequence that you want to fold. So this is the background on threading and it is very useful in cases where homology modeling fails specifically in cases where you have a low alignment and identity. This is also known as the twilight zone. So looking at the inputs and outputs of threading you will see that we have two structurally similar proteins and in both of them we have a fold that is common and this contains i, j, k and l secondary structures and we also measure their spatial adjacencies that is how close is i to j and l to k and so on. So the spatial adjacencies they describe the interactions between the secondary structures and then if given a sequence like that for which you don't know the structure then you can compare portions of this structure to the secondary structures and declare this portion to be taking up this specific structure. Similarly this portion will take up j and this portion will take up k and lastly this portion of the sequence will take up the secondary structure l. Now in this way you have predicted the structure for this sequence. So coming back to the issue of advantages and disadvantages. So the first thing is that threading or fold recognition is an extremely useful strategy for predicting the structure, the tertiary structure of proteins. On one side we have the homology modeling and on the other side is ab initio modeling. So threading is a compromise between the two. In case of homology modeling we are relying heavily on the template. While in ab initio we are relying on energy minimization which does not have anything to do with the template. But for fold recognition we are trying to find out the optimal secondary structures only or folds which are best suited to your sequence. So in this way you can create mix and match from the different folds that can be taken as structural elements of the sequence that you are considering. So as I just mentioned for the twilight zone in the alignment versus identity graph if there is low alignment and low identity you can consider that to be the twilight zone. So in this twilight zone you only have two options. You cannot use homology modeling of course. So the two options that you have is fold recognition or threading versus ab initio. So from these two of course threading is the simpler one and is more accurate as well. Towards the disadvantages one can obviously think that if you are trying to take some folds from one protein, some other folds from some other protein and in this way creating a combination of these folds towards creating a new protein structure then there is a high chance that the structure that you will create may not be biologically plausible. So the validation of the structures that are output from threading or fold recognition is essential. The result of such a shortcoming is that fewer than 30% of the predicted structures have homologs. So therefore you end up creating such structures by using fold recognition which do not have homologs in the PDB. So this has to be taken care of by a process called validation. So this is the major disadvantage of fold recognition.