Homology modeling
Mandatory reading
- https://en.wikipedia.org/wiki/Homology_modeling
- https://bioinfo.se/wp-content/uploads/2017/01/Bioinformatics-2005-Söding-951-60_1_.pdf
- https://bioinfo.se/wp-content/uploads/2017/01/journal.pcbi_.1000151.pdf
Slides
Videos
- https://youtu.be/k5jtmqYj6rs
- How to use homology modeling
- Principles of homology modelling
- Loop modelling
- Errors in homology modelling
- Modelling using swiss-modeller
- Swiss Modeller
- Modeller
- Modelling using modeller
- CASP
- Homology modelling benchmark
- Model Quality Assessment using ProQ
The link for Modelling using Modeller isn’t right ( It links to a python page)
In “principle of homology”, on an early slide it says using FASTA, or BLAST to search for homologs. Would you mind explaining how? And would it not be better to use multiple sequencing alignment at this stage?? e.g. using PSI-BLAST? And what is the reason for modelling loops first between side chain? And is minimisation referring to most thermodynamically stable form? And how many times do you generally iterate it? Is it until the most stable form has been found (if so, how do you determine that?)
Please explain the slide “the crucial importance of the alignment” and “Improving the alignment” in “principle of homology” again. ( Couldn’t hear it clearly)
How do you search for the fragments for the loops in the database ? Please explain again. I can’t hear clearly what the reason is for the break down of the method when the loops larger than 9, could you explain again. For the fragment library, how do you build it?
What is side chain packing?
Please elaborate the last slide ( checks of the structure) in homology modelling benchmark video again. Can’t hear clearly
On slide “input parameters” in the ProQ video, you mentioned the problem about using homology modeling( and Ramachandran plot ?!) is because it uses some residues. ( I tried to replay at least 3 times and I still couldn’t hear you, would you mind telling us what those residues you are talking about? )
THX.
1. In the video “Principle of homology modeling”, how would you “shift” the model in the case of duplication to avoid bulging? I am not very clear.
2. In video “HM benchmark”, could you give an example of “bad residue” and “bad chemistry” respectively?
3. Could you elaborate more on the slide about different strategies we employ for different sequence similarity?
The modelling with modeller video has been updated.