BLAST run - assigning which one as query and subject?

Question

[pbigbig]

Hi everyone,

I have some confusion here that would be great if you could clarify: Supposed I have a large set of scaffold in fasta format (from a genome assembly for example, and it may contain assembly errors), and I have a small reference cDNA set (obtained from Ensembl, so it can be considered as high quality reference). Normally, I was told that the larger set should be the subject for BLAST-ing and the smaller one should be used as query. Thus, should I makeblastdb of my large scaffold set and query the reference cDNA set against it? or doing vice-versa? (I have the feeling that using ref cDNA set as query is quite counter-intuitive because its role is for reference, so it should be the subject for BLAST-ing, isn't it?)

Thank you very much for any suggestion and clarification!

Answer 1

[Michael]

Using the small sequence set as a reference would give false low E-values, as those depend on the DB size and not on the total length or number of query sequences tested (Run a certain query sequence for itself, and with a bunch of other queries, the E-values are identical). You would get a massive multiple testing problem which is not corrected for. The assembly is much closer to the complete reference of all sequences that occur in this organism.

[pbigbig]

Hi Michael, thank you very much for your clarification.