@@annamhussainofficial2535 But as I know Alphafold structure is also a AI Structure, not the real one. Why would I predict model based on a Predicted AI structure! please let me know.
Yes exactly. But you can use template structures of your own sequence if you have. It will build structure of your given sequence by using your template.
@@AdityaKamde-x4j t1 is the template one first you have to download template from which is in pdb format then place it in your folder and after running the command it will be generate automatically. It is not already present it is the output file.
aoa mam I have mad my protein model by swissmodel by the help of your previous video. now my structure has been validated by ERRAT (100%) AND ramachandran (0 error) and but it is not being proved by verify 3d even though i haved used graly refine .. even in molbrobility it has 7 green parameters and 3 yellow. Can i use the structure for docking?
@@annamhussainofficial2535 JAZAKALAH MAM. AND KINDLY KEEP MAKING SUCH VIDEO AS WITH THE HELP OF UR VIDEOS I AM ABLE TO DO INSILICO ANALYSIS FOR MY MPHIL PROJECT
@@prareetamahapatra6660 oh well if so than copy the ready made Ali file from example folder and replace your own sequence in it. But remeber you have to delete the revious sequence and replace ot with your own sequence and than paste your sequence here rest will be same
@@annamhussainofficial2535 i am confused at one point you said prefer to take a structure that has been validated by experimental studies and the method for my template is alphafold.... kindly clear this
@@hibabukhary-q4f for homology modeling if you have more than 70℅ identity than you can confidently go for swissmodel in order to predict your structure but if its not. than you have to swtich at other ab initio methods like itasser and than these structures need to be validate also becaz their quility is poor and rough so we have to enhance its quality. And you can also enhance the quality of those structures which are predicted through homology modeling if its quality is poor.
Sure a lot of contents specifically at structure modelling is being prepared and will be uploaded very soon with excellent graphics. Actually i try to put a lot or at least maximum effort from my side before uploading any lecture which takes time as i m also a full time university teacher also. Only barrier for delay is just the time management but don't worry i m trying hard and will upload its new lectures very soon
Mam in which condition we prepare model by Homolgy modeling if the structure is already present in PDB database then we can use it or we will prepare it by modeling
@@wasifullahdawar6962 if its already available at PDB than you dont need to predict it from any method at all. PDB is the best database who predict perfect structures already.
Assalam o alikm teacher I build a protein structure on wincoot after that refined on phenix but still it's validity is not increasing more than 70%,even i minimize it many times on chimera please give any suggestion
mam how do we find sequence which don't have structure and then how we find the template which have 30 plus similarity please tell me asap and I am from India ur other videos help me a lot thank. u for that please resolve my query
@@hhh-mp8kw for structure please check in pdb and other structural databases if the structure is predicted than you will get there but in case the structure is unpredicted than you could not find in any database.
the maximum sequence length for protein modeling on Swiss-Model was around 1500 to 2000 amino acids you can use modeller for large sequence length as MODELLER typically doesn't have a strict fixed limit on sequence length, but larger proteins may require more computational resources and time for modeling.
The link for this sequence is here: www.uniprot.org/uniprotkb/P03954/entry#sequences now the interface of UniProt has changed completely. now we can either copy and download but before two years we can get FASTA sequence in this form as it is shown in the video
You can use "swiss model" if they have any template with similarity of at least 70%. otherwise you have to use "itasser" For any structure which dont have any sufficient similarity of at least 70℅ with a template( any other predicted protein structure at pdb or any other database).
Thank you for this fantastic video, ma'am. I've understood the whole concept, but my device has some issues. Whenever I try creating an NJ tree for the aligned sequences, MEGA becomes unresponsive, and the application closes on its own. Could you please help me with this?
Thank you Ya actually your mega software is outdated and you need to reinstall it. First uninstall it and then reinstall it then hopefully you issue will be resolved.
Using modeller is anther option for doing homology modeling but it will do manually for you.. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-rD8XcJaZanw.html
if you can not open don't worry. You can copy the Python script from the basic example from modeler original website page salilab.org/modeller/tutorial/basic.html and past it in some text or notepad file and make it by your own .. simple. or just use my other method which is a shortcut method for using modeller by using this link... ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-2j5WV5MEWt4.html
Thanks.. Actually instructions are quite clear but its natural that when students learn modeller in start they feel that its confusing and a little vit difficult but as they become expert of it then they don't get and feel difficult anymore.
