There are at least three ways of sequence numbering for proteins with signal peptides or other types of N-terminally located presequences: 1) to start numbering from first translated residue; 2) to start numbering from the first residue after signal peptide; 3) to use a combined numbering by showing both numbers In my view, approach #3 is more preferable, since it clearly eliminates potential errors in interpretation.

Hi there,

Numbering is associated to the translation product. The main interest is that it tells you about the story of the mature protein. For instance, to understand the maturation process and structure of chymotrypsine the numbering is very useful:

https://books.google.fr/books?id=2yRDWkHhN9QC&pg=PA118&lpg=PA118&dq=chymotrypsinogen+maturation&source=bl&ots=mQnMo3aZX0&sig=55twrlMgBT9SQVsrw8lbxSFTvDc&hl=fr&sa=X&ved=0ahUKEwjI5PHcyJ7UAhXSJVAKHZDiBVsQ6AEILTAB#v=onepage&q=chymotrypsinogen%20maturation&f=false

I agree with Kou and Dominique. Normally it is the from the start of translation. The reason is can be confusing is that when you are working from translation of a DNA sequence you can see everything, but in older literature they might have discussed the final protein since the data was protein based. So you will see both being used in the literature.  BUT the most important this is to specify when you are writing, what convention you are using. 

I totally agree with Benedik. Although in many publications they refer to the numbering of the complete genome, numbering from the beginning of the translation makes the analysis easier when studying a mutation for example. By all means, it is imperative that the author makes his method clear. Even more if the analysis wasn't performed in the intire genome sequence.

There are at least three ways of sequence numbering for proteins with signal peptides or other types of N-terminally located presequences: 1) to start numbering from first translated residue; 2) to start numbering from the first residue after signal peptide; 3) to use a combined numbering by showing both numbers In my view, approach #3 is more preferable, since it clearly eliminates potential errors in interpretation.