I downloaded a PDB structure of a protein and the FASTA sequence of the same protein. When I imported the structure in PyMol, the sequence represented in one subunit of a protein is only a portion of the entire sequence mentioned in FASTA sequence. I am confused about this entire thing. Any help will be highly appreciated.
Dear Debanjan Kundu ,
Without an example it is somewhat difficult to pinpoint the answer. However, it can be because of a number of reasons:
-The most likely one is that you need to realize that a structure is most of the times based on crystallographic data. It is quite common that parts of the protein are not included in the structure (not enough details to extract the structure of that part or because taking a smaller part of the protein leaves better crystal qualities).
-Another reason might be that the structure of a protein is taken from a crystal structure from a complex of proteins (like for example SecYEG, which is already a complex, with or without SecA) and ‘isolating’ one of the proteins might give some differences.
-Sometimes proteins form (in crystals) multicomplexes (dimers or higher).
Hope this clarifies it somewhat.
Best regards.
The sequence you download should be the sequence of the construct that was used for crystallization, although that seems to be a rather loose rule, as some authors include parts of the sequence that are cleaved off during sample preparation, e.g. signal peptides, or submit the entire UniProt sequence even if they crystallized only a fragment.
In addition to subsequences that never entered the structure determination, some parts of the molecule simply are not resolved well enough to be seen in the experimental structure. Especially if the structure is not an X-ray, but a cryo-EM structure, there may be large domains missing because they were not sufficiently well ordered to be seen in the structure. Check the associated publication (if available) to see how the authors explain the discrepancy.
Martiniano Bello, the UniProtKB entry associated with a given chain in a protein structure may be even farther from the sequence embodied by the structure.
UniProtKB gives you the sequence of the biological protein as encoded on the natural mRNA. However, to produce a protein in sufficient amounts for structure determination, you frequently have to significantly change that sequence. You may need to get rid of parts of the protein that are predicted to be natively unfolded, add foreign domains that improve expression and/or might help crystallization (e.g. MBP fusions), introduce mutations that improve folding and stability, add tag sequences to help purification and detection, replace flexible loops by shorter ones, or even by folded domains (see GPCR structures), introduce specific protease cleavage sites to remove parts of the protein needed only for expression and purification.
In the SEQRES part of the pdb header, which is where the fastA file you can download from the pdb comes from, you want to find the theoretical sequence of the protein as it was entered into the structure determination process. This is what you compare your query sequence to if you do a blast search of the protein structure database.
You may want to compare this theoretical sequence of the construct to the UniProtKB record as well as to the sequence you extract from the coordinates. However, what Debanjan Kundu asked was why some parts of the protein theoretically present in the structure do not have coordinates attached.
Thank you one and all for your explanations, I shall keep this thread existing for any further queries Annemarie Honegger , Rob Keller and Martiniano Bello !
In multi-chain structures, also check out the other chains for the same pdb entry - I have seen a few instances where the sequences appear to have been swapped.
Also, if we check the structure and FASTA sequence of PDB_ID (1G6N) then the preceding "MVLGKPQ" might be the signal sequence facilitating the protein translation or protein purification effort. What I say is there may be some signalling sequences in the N-terminal region of the sequence.
FASTA sequence: MVLGKPQTDPTLEWFLSHCHIHKYPSKSTLIHQGEKAETLYYIVKGSVAVLIKDEEGKEMILSYLNQGDFIGELGLFEEGQERSAWVRAKTACEVAEISYKKFRQLIQVNPDILMRLSAQMARRLQVTSEKVGNLAFLDVTGRIAQTLLNLAKQPDAMTHPDGMQIKITRQEIGQIVGCSRETVGRILKMLEDQNLISAHGKTIVVYGTR
Sequence extracted from structure:
TDPTLEWFLSHCHIHKYPSKSTLIHQGEKAETLYYIVKGSVAVLIKDEEGKEMILSYLNQGDFIGELGLFEEGQERSAWVRAKTACEVAEISYKKFRQLIQVNPDILMRLSAQMARRLQVTSEKVGNLAFLDVTGRIAQTLLNLAKQPDAMTHPDGMQIKITRQEIGQIVGCSRETVGRILKMLEDQNLISAHGKTIVVY
This "missing" residues may also be present in the crystallised construct, but are too disordered in the crystal to be visible in the electron density, e.g. the C-terminal GTR. They therefore do not appear in the coordinates. In addition. sometimes if only the main chain coordinates can be assigned, these residues may be called "Ala" although they are different residues.
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Biomolecules
- Peptides