Makeblastdb from paired end reads

[sp24]

Hello everyone. I’m in the process of analyzing some transcriptome data. I’m not quantifying expression at all. I only want to search for sequences based on a query file of my genes of interest. I have paired end reads, so I’m not using a fully assembled transcriptome. Basically, I need some advice since I’m new to this area.

What I have done so far is merge the R1/R2 files and convert the files from Fastq to Fasta using the FastX-toolkit. I used the makeblastdb command to make the blast database from the resulting Fasta file. I know that I’m supposed to get .nhr .nin and .nsq, but I think that the database is so big that I got something like this:
DB.00.nsq, DB.00.nin, DB.00.nhr
DB.01.nsq, DB.01.nin, DB.01.nhr
and so on.
So here’s the first question: is this a problem? Or will I just have to blast my query file against each database (00, 01), one at a time?

Also, before I get too far into this, I also would like to know if for some reason I shouldn’t be merging the read files and creating a database from it.

Thank you for taking the time to read this!

[maubp]

I know that I’m supposed to get .nhr .nin and .nsq, but I think that the database is so big that I got something like this:
DB.00.nsq, DB.00.nin, DB.00.nhr
DB.01.nsq, DB.01.nin, DB.01.nhr
and so on.
So here’s the first question: is this a problem? Or will I just have to blast my query file against each database (00, 01), one at a time?

No that's normal for a very large database - have a look at the NCBI provided NR or NT databases as an example.

Also, before I get too far into this, I also would like to know if for some reason I shouldn’t be merging the read files and creating a database from it.

That is a very sensible question - you might get something out of your planned analysis but this is not the normal approach (I would do a transcriptome assembly giving you putative transcripts, attempt to analyse them, for example with BLAST against sister species).

[sp24]

Thanks for your response. I decided not to merge the R1/R2 files since I've been reading on here that the R2 file needs to be reverse complemented. I've been creating blast databases with individual read files and blasting against my query, but now I'm trying to figure out how to interpret the outputs.

For the scope of this project I'm not going to be able to assemble transcriptomes, just trying to figure out how to retrieve these sequences. If someone reads this and has done this before, I would appreciate your input. Or if there's any papers/online resources you can point me to that would be great as well.