I'm working on my first sequencing related project and I'm trying to find proteins with a specific PFAM ID (PF11999). The project is called "MMETSP", I searched the annotations for that ID, identified signal peptides using SignalP and their target locations with a tool called MultiLoc2 (I was looking for extracellular targets only).

Out of the very few sequences that where left after all this filtering, none of the DNA-bits started with the basepair-code "ATG", for Methionine. How can this be?

I used the shell to calculate that only 1,83% of all the sequences start with ATG.
Any ideas on this?

  • 1
    $\begingroup$ Exactly what sequences are you searching? $\endgroup$ – Devon Ryan Jul 28 '16 at 12:43
  • $\begingroup$ I'm looking for Antifreeze proteins (PF11999) in marine diatoms, that are located in the extracellular region. Using the cleavage site provided by SignalP I can isolate the targeting peptide. However, these peptides seem to have no "start"-code. I got the sequences by looking up the protein IDs in the fasta files that I got from the annotation-search. $\endgroup$ – Birg3r Jul 28 '16 at 14:17
  • 2
    $\begingroup$ You're just looking at the cleaved peptide sequence then? There's no reason those need to start with a methionine unless they're type-I and don't receive further N-terminal processing. $\endgroup$ – Devon Ryan Jul 28 '16 at 18:57
  • $\begingroup$ Well, I'm looking at the DNA coding the peptide. The peptide is located in front of the "maternal" protein, so the ORF should start with a methionine, shouldn't it? $\endgroup$ – Birg3r Jul 29 '16 at 8:28
  • 1
    $\begingroup$ If there are no UTRs in this species and the CDS is annotated as full length then yes. Otherwise no. $\endgroup$ – Devon Ryan Jul 29 '16 at 12:33

If anyone stumbles across this question, I found out what the problem was in the end:
contrarily to my assumption that the DNA sequences where all in proper 5'3'-direction, it turned out that we found the exact ORF on the complementary strand in the other direction and who knows where else. Luckily, the MMETSP-Data also featured a /pep directory, in which the needed sequences where found in a clean version. Thanks for reading and good Luck.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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