why we need the reads?

[huanggm]

hi,
I have a question,in the paired-end-read method,we will get two paired reads,why we need the reads, and how can we use the use?what does them used for?

[DFJ111]

Two reasons
1) You get more sequence
2) The two reads are separated by a region where you do not know the sequence, but you do know the distance between the reads. Genome assembly programs can use this information.

[ymc]

Is it possible to specify the insert size for bwa? It seems to me we can't and therefore we can't map small number of reads, right??

[dpryan]

Is it possible to specify the insert size for bwa? It seems to me we can't and therefore we can't map small number of reads, right??

You can specify the maximum insert size with the "-a" option. I guess if there were truly very few reads then you could output a bunch of matches and write a small script to just select the most likely. Having said that, most aligners are designed to handle large numbers of reads, so you might be better off using different tools for very few reads.

[azneto]

Bowtie allows you to specify both minimum and maximum insert size. Assemblers usually have the parameters min_ins, avg_ins, and max_ins.

[swbarnes2]

bwa aligns the reads separately, you don't need to tell it your estimated insert size.

If one end reads, and the other doesn't, it will look at the other reads from your sample, work out the likely range of insert sizes, and will try harder to make the second one map in a sensible location.

But in general, bwa will just align them both independantly.