compare solexa and 454

[qinlian]

hi gus,
Has anyone here compared one run data between solexa and 454?
It seems that one run of solexa can gernerate about 50000000 reads, 35 bp average, while a complete run of 454 (GS FLX ) can have about 400000 reads with average 250 bp.
One gets longer, while the other gets more.
I am confused now, i don't know which one is better for de novo assemble (there's no reference sequence)?
does anyone have the experience?

[bioinfosm]

It depends on a lot of things - what kind of genome is it, how big is it, repeats, etc?

For smaller simpler genomes Solexa might be the choice, but otherwise, 454's longer reads will give a much better assembly. Another point to consider is, do you care to assemble only the informative (non-repeat) regions; and then as you said, coverage will be different if you use a single run of one platform... and also the cost.
so, essentially, its your call ... unless you have more details here

[ScottC]

I think the best option is a run of each, if you can afford it! Then you get the depth and homopolymer coverage from Illumina and the repeat coverage from 454. As bioinfosm stated, it also depends on what kind of sequencing project you're doing (i.e. do you just want the majority of the sequence, or do you want a complete, closed genome).

Cheers,

Scott.

[qinlian]

Thanks for u reply!
We plane to do an eukaryote transcriptome sequencing, and we hope to get more coding region of coding genes.

[sparks]

You might also consider using paired end reads especially with Illumina.With Illumina paired end sequencing you'll get +30bp at each end of a 250bp fragment. This will help a lot with assembly.
454 paired end offers even longer insert lengths and should help identify splice forms.