MS Amanda stores databases internally?

[schulzes...@gmail.com]

It seems like MS Amanda stores databases internally, using the filename as an identifier.

This can lead to some unexpected results, when not being aware of this feature.

In my case, I was running a search on file x (Trypsin digest) against database1 (/user/databases/trypsin/example_database.fasta). This is a target decoy database, where the decoys have been generated by peptide shuffling (taking into account trypsin cleavage sites).

After that, I was running a search on file y (GluC digest) against database2 (/user/databases/gluc/example_database.fasta). This is the same database, but decoys have been generated by peptide shuffling based on GluC cleavage sites instead of Trypsin cleavage sites.

In this case, (decoy) peptides were identified that are not part of database2, but part of database1 (part of decoy proteins based on Trypsin cleavage sites).

If I rename database2 (/user/databases/gluc/example_database_gluc.fasta) this problem does not occur.

I am aware that it is not ideal to name different databases the same way.

However, in my defense, they have a different path and the full path is used in the database settings. So the settings.xml differs (/user/databases/gluc/example_database.fasta) from what is actually used (/user/databases/trypsin/example_database.fasta).

So if my assumption that MS Amanda stores the database internally is correct, it would be nice if this would be mentioned in the documentation and maybe the identifier that is used should not only be the filename but the full path? I guess the latter is debatable, but I just wanted to make aware of this "issue".

The version that I have been using is the most recent standalone MS Amanda 2.0 v. 2.0.0.11219 on a Linux as well as Mac system.

[Viktoria Dorfer]

Hi Stefan (?),

this is indeed true. MS Amanda performs an enzymatic digestion of the database file only once per setting (being enzyme, cleavage site) for performance reasons. MS Amanda checks whether it finds a digest with the same name, same settings and the date when the file has been changed the last time (this is to avoid missing updates of the same database file).

Taking the full path as identifier is difficult as at some point the filename will be too long, but we will think of a solution for this issue, breaking down the full path to a short identifier.

Thanks for reporting this!

Best regards,

Viktoria

[Viktoria Dorfer]

Hi,

the new MS Amanda standalone versions now enables to have the same database name if they are in different folders.

Best regards,

Viktoria