BWA-MAQ questions?

[baohua100]

We have RNA-Seq paired-end data, the insert size is about 200bp , and the read length is different.

4_1.fq 50bp
4_2.fq 36bp

5_1.fq 75bp
5_2.fq 50bp

Can i run

bwa aln database.fasta 4_1.fq 4_2.fq 5_1.fq 5_2.fq > aln_sa.sai


How BWA do paired-end alignment for different read length?

And there are many situations like : one read is in one exon ,the other read is in other exon, the mapping position of these two reads on genome is 500, 3600. The distance is 3100 (there is a intron there) and BWA will discard these paired-end reads alignment because their distance is signicicant larger than 200bp?

Or i should map these reads as single reads?

[totalnew]

Firstly, you will have to align fastq file individually to get correspond suffix array files, as

bwa aln database.fasta 4_1.fq > 4_1.fq.sai
bwa aln database.fasta 4_2.fq > 4_2.fq.sai
bwa aln database.fasta 5_1.fq > 5_1.fq.sai
bwa aln database.fasta 5_2.fq > 5_2.fq.sai

Then use sampe to generate results in sam format, I never align PE reads with different length, but I guess you may be able to get result sam files, just displaying different length at SEQ and QUAL fields. (Try and let us know what you get)

I have no idea about the insert size issue, as bwa is a new tool, everyone is doing experiements on that.