Hi everyone,
I am searching for all PDB structures that have been solved for a specific protein. There are many homologs of this protein so I would like to narrow my search to structures that have 100% sequence identity. I have the a protein sequence, name, and a few PDB of the structure by it self and in complex with DNA. Does anyone had any tools or trick they use to get a complete list of all structures for a particular protein. Any help will be appreciated!
Thank you in advance
rcsb.org can help with that. You can search by sequence similarity and limit your results to 100% sequence identity. See the example for the alpha chain of hemoglobin (4hhb.A) below. You can replace the sequence with whatever you're interested in. If there's a PDB structure for your protein of interest, you can type that into the "Entry ID" field and it will fetch the sequence for you. You can also adjust the identity cutoff value as needed.
https://www.rcsb.org/search?request=%7B%22query%22%3A%7B%22type%22%3A%22group%22%2C%22nodes%22%3A%5B%7B%22type%22%3A%22terminal%22%2C%22service%22%3A%22sequence%22%2C%22parameters%22%3A%7B%22value%22%3A%22VLSPADKTNVKAAWGKVGAHAGEYGAEALERMFLSFPTTKTYFPHFDLSHGSAQVKGHGKKVADALTNAVAHVDDMPNALSALSDLHAHKLRVDPVNFKLLSHCLLVTLAAHLPAEFTPAVHASLDKFLASVSTVLTSKYR%22%2C%22identity_cutoff%22%3A1%2C%22sequence_type%22%3A%22protein%22%2C%22evalue_cutoff%22%3A0.1%7D%7D%5D%2C%22logical_operator%22%3A%22and%22%7D%2C%22return_type%22%3A%22polymer_entity%22%2C%22request_options%22%3A%7B%22paginate%22%3A%7B%22start%22%3A0%2C%22rows%22%3A25%7D%2C%22results_content_type%22%3A%5B%22experimental%22%5D%2C%22sort%22%3A%5B%7B%22sort_by%22%3A%22score%22%2C%22direction%22%3A%22desc%22%7D%5D%2C%22scoring_strategy%22%3A%22combined%22%7D%2C%22request_info%22%3A%7B%22query_id%22%3A%22aeaac5009f6dfc83470b36456129d46b%22%7D%7D
Hi, I-Tesser server may help you. You can paste your protein sequence (in Fasta format) and I-Tesser will predict the structure of your protein of interest (5 predictions). You will also have values about the most consistent prediction, as well as the most closer 3D reconstruction in PDB altready availables (by cristallography). Once you get the good prediction, you could "modify" your protein of interset by using DeepView software.
You may also use AlphaFold.
Good Luck
P
The problem with insisting on 100% sequence identity is that you will miss any sequences that have been slightly altered to improve protein production, stability or crystallisation (e.g. tags added to aid detection and purification, stabilising mutation, mutations in an enzyme active site to allow co-crystallization with a substrate by blocking turn-over, removal of glycosylation sites to obtain a more homogeneous product). In addition, for larger proteins you may need to find different structures of partial constructs. Once you find a a single structure for your target sequence on https://www.rcsb.org , you can directly access all structures that map to the same UniProt record: see image
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics
- Bioinformatic Tools