2
$\begingroup$

I wanted to compare the amino acid sequence of enzymes for this project I'm working on and need to compare them at their catalytic site. For that, I went to the Catalitic Site Atlas to get the information on the catalytic site, but since they don't offer an easy way for me to download the structure data programmatically I downloaded it from the RSCB PDB by downloading the fasta sequence. When checking for the catalytic sites it wasn't matching what the CSA was telling me and that's when I realized that they are different files. Take for example the 3nos, the CSA presents the following sequence:

MGNLKS...

While the PDB presents the following sequence:

PKFPRV...

Why aren't they the same sequence if it's the same protein?

Sorry if it's a noob question, I'm not a biologist, just a computer scientist who happens to like bioinformatics.

Important info:

The CSA data comes from here while the PDB data comes from here

Improve this question
$\endgroup$
2
  • $\begingroup$ Please edit your question and give us the exact links you used to download. The simplest explanation is that you are not actually looking at the same protein, but we can't be sure unless you show exactly what you are comparing. For instance, your RCSB link shows two sequences but no information about what species they are from, what isoform, nor whether they're the entire protein. In fact, they look like protein fragments (most proteins start with an M). My guess is that you are looking at the entire sequence in one site, and a subsequence in the other but I can't be sure. $\endgroup$
    – terdon
    Oct 15, 2017 at 14:53
  • $\begingroup$ There you go, sorry about that $\endgroup$
    – João Areias
    Oct 15, 2017 at 16:02

3 Answers 3

Reset to default
3
$\begingroup$

Crystallography results (pdb files) almost always contain a truncated sequence.

Both ends of a protein are often flexible (even in a crystal) and don't result in enough data for a good fit. The corresponding residues are removed from the model and the sequence, and you're left with only the residues that show a defined electron density.

Improve this answer
$\endgroup$
0
1
$\begingroup$

One sequence is partly contained in the other (highlighted).

So the CSA sequence is (FASTA format, truncated):

>sp|P29474|NOS3_HUMAN Nitric oxide synthase, endothelial OS=Homo sapiens GN=NOS3 PE=1 SV=3
MGNLKSVAQEPGPPCGLGLGLGLGLCGKQGPATPAPEPSRAPASLLPPAPEHSPPSSPLT QPPEGPKFPRVKNWEVGSITYDTLSAQAQQDGPCTPRRCLGSLVFPRKLQGRPSPGPPAP EQLLSQARDFINQYYSSIKRSGSQAHEQRLQEVEAEVAATGTYQLRESELVFGAKQAWRN...

taken from http://www.uniprot.org/uniprot/P29474 for convenience.

While the PDB one is :

>3NOS:A|PDBID|CHAIN|SEQUENCE PKFPRVKNWEVGSITYDTLSAQAQQDGPCTPRRCLGSLVFPRKLQGRPSPGPPAPEQLLSQARDFINQYYSSIKRSGSQA HEQRLQEVEAEVAATGTYQLRESELVFGAKQAWRNAPRCVGRIQWGKLQVFDARDCRSAQEMFTYICNHIKYATNRGNLR SAITVFPQRCPGRGDFRIWNSQLVRYAGYRQQDGSVRGDPANVEITELCIQHGWTPGNGRFDVLPLLLQAPDEPPELFLL...

The Uniprot entry mentions 3 different isoforms due to alternative splicing, so perhaps that is what is going on here. Here is the output from a sequence alignment (using https://www.ebi.ac.uk/Tools/psa/emboss_matcher/):

#=======================================
#
# Aligned_sequences: 2
# 1: NOS3_HUMAN
# 2: SEQUENCE
# Matrix: EBLOSUM62
# Gap_penalty: 14\
# Extend_penalty: 4
#
# Length: 240
# Identity:     240/240 (100.0%)
# Similarity:   240/240 (100.0%)
# Gaps:           0/240 ( 0.0%)
# Score: 1294
# 
#
#=======================================

NOS3_HUMAN        66 PKFPRVKNWEVGSITYDTLSAQAQQDGPCTPRRCLGSLVFPRKLQGRPSP    115
                     ||||||||||||||||||||||||||||||||||||||||||||||||||
SEQUENCE           1 PKFPRVKNWEVGSITYDTLSAQAQQDGPCTPRRCLGSLVFPRKLQGRPSP     50

NOS3_HUMAN       116 GPPAPEQLLSQARDFINQYYSSIKRSGSQAHEQRLQEVEAEVAATGTYQL    165
                     ||||||||||||||||||||||||||||||||||||||||||||||||||
SEQUENCE          51 GPPAPEQLLSQARDFINQYYSSIKRSGSQAHEQRLQEVEAEVAATGTYQL    100

NOS3_HUMAN       166 RESELVFGAKQAWRNAPRCVGRIQWGKLQVFDARDCRSAQEMFTYICNHI    215
                     ||||||||||||||||||||||||||||||||||||||||||||||||||
SEQUENCE         101 RESELVFGAKQAWRNAPRCVGRIQWGKLQVFDARDCRSAQEMFTYICNHI    150

NOS3_HUMAN       216 KYATNRGNLRSAITVFPQRCPGRGDFRIWNSQLVRYAGYRQQDGSVRGDP    265
                     ||||||||||||||||||||||||||||||||||||||||||||||||||
SEQUENCE         151 KYATNRGNLRSAITVFPQRCPGRGDFRIWNSQLVRYAGYRQQDGSVRGDP    200

NOS3_HUMAN       266 ANVEITELCIQHGWTPGNGRFDVLPLLLQAPDEPPELFLL    305
                     ||||||||||||||||||||||||||||||||||||||||
SEQUENCE         201 ANVEITELCIQHGWTPGNGRFDVLPLLLQAPDEPPELFLL    240
Improve this answer
$\endgroup$
1
$\begingroup$

This answer is correct, I just wanted to add that the correct sequence numbering is preserved in the PDB file in the DBREF record (which you can see by opening the PDB in a text editor):

DBREF 3NOS A 66 492 UNP P29474 NOS3_HUMAN 66 492

In plain English, the sequence presented in this file (3NOS chain A) corresponds to residues 66 - 492 of the associated UniProt (UNP) entry (accession: P29474).

Improve this answer
$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .