Blast or mapper ?

Question

[Picasa]

Hello,

What are the advantages of using a homology research (such as blast) over a mapper (bwa, bowtie) ??

[h.mon]

Blast is a similarity search, not homology search. And if you do not include your question to narrow it down to a specif problem, you will get all sorts of answers.

[abascalfederico]

It is both, similarity and homology search. The success of BLAST relies not only on its speed but on its better model to assess statistical significance of similarities, i.e. whether a similarity (score) can be reached by chance or a common ancestry (homology) can be inferred.

[Picasa]

I'm looking to classify me sequence (find which taxa are my sequences)

Answer 1

[WouterDeCoster]

That entirely depends on the application/purpose of course!

[Picasa]

Indeed ... For example, if I'm looking to find from which species are my sequences ?

[GenoMax]

Either will work for that application. In fact, blasting a few reads at NCBI is the diagnostic tool of choice to answer your question (if you suspect that the data may be contaminated).

[Picasa]

Ok, but what the advantages of using blast ?

[abascalfederico]

Mappers are much more strict, they expect/look for near-exact matches between the reads and the genome. In contrast, BLAST allows you to look for distant homologies. Another difference is that with BLAST you can scan all known species in one run, whereas mappers expect one single genome.

[Devon Ryan]

It has a nice web application that allows you to just paste a few reads into and click a button. Also, you're going to have to spend a LOT of time to make an index of the size available from NCBI's public blast server.

Answer 2

[Antonio R. Franco]

Bowtie and BWA are designed for a different purpose. To align billions of reads to a reference genome or transcriptome To accomplish that, they index the reference, and then, reads are mapped very rapidly and efficiently

BLAST will need days, weeks or even more to do the same task