6
$\begingroup$

I am wondering about the appropriate workflow to determine a protein's cellular location based on its sequence.

Let's say I have a sequence like this from a fasta file

MTIKEHKVVYEAHNVKALKAPQHFYNSQPGKGYVTDMQHYQEMYQQSINEPEKFFDKMAK
EYLHWDAPYTKVQSGSLNNGDVAWFLNGKLNASYNCVDRHAFANPDKPALIYEADDESDN
KIITFGELLRKVSQIAGVLKSWGVKKGDTVAIYLPMIPEAVIAMLAVARIGAIHSVVFAG
FSAGSLKDRVVDANSKVVITCDEGKRGGKTINTKKIVDEGLNGVDLVSRILVFQRTGTEG
IPMKAGRDYWWHEEAAKQRTYLPPVSCDAEDPLFLLYTSGSTGSPKGVVHTTGGYLLGAA
LTTRYVFDIHPEDVLFTAGDVGWITGHTYALYGPLTLGTASIIFESTPAYPDYGRYWRII
QRHKATHFYVAPTALRLIKRVGEAEIAKYDTSSLRVLGSVGEPISPDLWEWYHEKVGNKN
CVICDTMWQTESGSHLIAPLAGAVPTKPGSATVPFFGINACIIDPVTGVELEGNDVEGVL
AVKSPWPSMARSVWNHHDRYMDTYLKPYPGHYFTGDGAGRDHDGYYWIRGRVDDVVNVSG
HRLSTSEIEASISNHENVSEAAVVGIPDELTGQTVVAYVSLKDGYLQNNATEGDAEHITP
DNLRRELILQVRGEIGPFASPKTIILVRDLPRTRSGKIMRRVLRKVASNEAEQLGDLTTL
ANPEVVPAIISAVENQFFSQKKK

I can feed it to NCBI's blastp which then gives me a list of genes. I can check e.g. the first entry which also provides a link to SwissProt. There I finally see that this protein is located in the nucleus and also the cytosol.

My questions are:

1) Is there a more straightforward way of achieving this information?

2) Let's say someone annotated this protein to be a mitochondrial one. How could I check that i.e. (dis)prove that?

Improve this question
$\endgroup$
1
  • $\begingroup$ Any update on this? $\endgroup$
    – BND
    Aug 21, 2020 at 8:50

2 Answers 2

Reset to default
6
$\begingroup$

You could use Blast2go to find the predicted relevant gene ontologies of your sequences. If you look for the cellular compartment sub ontology you should find the location of such protein. I have never used this tool, but here is the paper describing how it works.

I wouldn't restrict to the first predicted match of blastp see if some more genes with a good fit are also located at the same sites. But otherwise your workflow seems right to me.

But to prove this is in X or Y you, or your team or collaborators, should always check with histochemistry if it really allocates on the predicted place. All the softwares predictions must be validated with some biological experiments to prove them.

Improve this answer
$\endgroup$
2
  • $\begingroup$ Thanks for the suggestion, I will give it a try once I find time. $\endgroup$
    – Cleb
    Nov 16, 2017 at 9:12
  • 1
    $\begingroup$ @Cleb I forgot to write that your worklow is how I would do it too. $\endgroup$
    – llrs
    Nov 16, 2017 at 9:21
2
$\begingroup$

You can run the sequence through tools like SignalP to look for cell surface signals and so on, which will give you a clue about where I is bound for.

None of these tools will give you an absolutely definitive answer though, often because we simply don’t have the experimental data.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Thanks, was not aware of this tool, will give it a try. $\endgroup$
    – Cleb
    Nov 16, 2017 at 18:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.