Illumina GAII paired end length differs

[cb4smit]

Hi
I recently received GA2 human paired end data. One thing that bothers me is that the length of both the reads combined is odd, 105bp, and not 72bp (2x36) or 100bp (2x50) or 120bp.

I asked the data provider about it, and she said that the cycles used during the run were 50 and 55.

Being a computer scientist by default, is it possible to have the "forward" and "reverse" read of different lengths? Presumable I can just split of the first 50bp to generate PE data for input into Mosaik/Velvet.

All help would be appreciated

Carl

[jnfass]

Hi Carl,

It's certainly possible to have unequal f/r read lengths ... the chemistry of the sequencing causes slow degradation of the quality of basecalls, but that depends on a lot of factors (reagent quality, planetary alignment, yada yada). Your sequencing provider may watch each run (the forward, and the reverse), and stop each once the average quality drops below an "acceptable" level, which could be at different cycles. Or, they may have very consistent runs, and have a policy of stopping at a certain cycle for every run. In any case, nothing about the sequencing process constrains the f/r read lengths to be equal.

I'm pretty sure velvet can actually handle unequal f/r lengths, but I'm not sure about mosaik.

~Joe

[cb4smit]

Hi Joe.
Thank you for the reply!