Database effective length on BLAST

[Galo]

Hello everybody,

First of all I did my homework and before ask, I looked for this question with google and in this forum but I didn't get the information that I want.
If I did wrong, I'm sorry.

My question is about appropriate use of database effective length option performing BLAST.

I have a scaffold of my specie of interest and I have to find whether a specific gene from another related specie is in this scaffold or not. The fact is that my scaffold has a lot of 'N' in such way that the 41.3% of it are 'N'.

I was wondering if it would be appropriate "tell" BLAST database effective length is a 41.3% smaller so I would get more fitted expected values.

For me it makes sense, but I am totally novice in these fields.

Thank you in advance!

[Galo]

So... Anyone?

[Galo]

Ok, I think I found the answer. I think that BLAST calculate database effective length without considering N nucleotides.

So, does anyone know what database effective length parameter is for?

[JamieHeather]

I don't know the answer, but there's an email option on the BLAST help page.

You could email them and ask (and then post the answer here, in case anyone else wonders in future!).

[Galo]

Thank you.

I sent them an e-mail just now and I ask them about use of Database effective length and x-drop values, wich I know what are they used for (I'm talking about x-drop), but I don't know how calculate or know proper values for our query.

I'll write the answer here!

[Galo]

I got the answer from NCBI.

They sent me a link where one download a PDF file about scoring in BLAST:
ftp://ftp.ncbi.nlm.nih.gov/blast/doc...er/scoring.pdf

About x-drop parameter they tell that there isn't a method to calculate the x-drop parameter but try-error method. Despite of the importance of x-drop parameter.