 Dear students, now we will continue building on our concepts in fold recognition. As you know fold recognition is also known as threading and it is a very important tool to predict protein structures. So this tool or technique is employed in case homology modeling which is the primary tool for predicting the protein structures does not work properly. So what happens in that case? We move towards the twilight zone where we have a low identity and a low alignment. So here let's work on an example to find the best fold for a sequence. So the best way to find that is to mount the residue sequence that is the primary sequence of a protein onto a known protein structure. So this known protein structure is coming from a structure database or fold database. One of such databases is called SCOP which means structural classification of proteins. So let's say if you find this structure in the SCOP database then what you do is you put your sequence along the backbone of this structure and you see how it matches with the fold from the database. So as you can see in this case this portion matches really nicely with the fold but this portion does not match with the fold. So such a situation can vary if you keep changing the fold. That is if you have if you keep bringing different fold structures then you can find different alignments between your sequence and your fold. Now once you have performed all such alignments you will find that fold from the database which matches your sequence perfectly. So in this case the circles they are the amino acids this is your fold structure. So you have mounted the amino acid sequences onto the fold structure. This is the fold from SCOP. Now you can do this for multiple structures of course, multiple folds of course and you can find that fold which best matches your sequence. So simply the process of threading is mounting of an amino acid sequence onto the backbone of a template fold. So the structures are folds. Next on each step we drag the sequence for instance MQVK LFTY. So if this is your fold then in the first step we will have MQVK LFTY and Y. In the second step M will move to the position of Q, Q will move to the position of V, V will move to the position of K and so on. So you will end up with MQVKL and so on. So M has moved from the first position in the fold to the second position. We keep shifting the sequence until we find the best match for the sequence in the fold. Next then for each comparison we compute the score. So the score or the fitness of the matching will determine at the end how good is the match. Typically Z score is used for this. So let's take an example. So this is a fold from the fold library. This is the amino acid sequence that you are going to fold onto the structure and you have a scoring function which is usually a Z score. Then you input these three things into your threading software and the output is a structure that is like this. So in this we have the fold that is here and some loops that are joining the different regions within this fold. So this is the overview the input and the output. In conclusion we pass the sequence over the various folds during threading and find out which sequence best matches the fold and we create a combination of those folds to create the tertiary structure and the best match is computed using a scoring function which is typically the Z score.