I think the critical factor is that at a cluster density where you produce the most sequence (even the most Q30 bases), the quality of single-plex amplicon runs deteriorates badly over the last 50 bases of a 250 cycle read. Whereas if you keep the density much lower--such that you may only generate 10 million or fewer clusters--the quality can be high all the way to the ends of the reads.
At these lower cluster densities we see pair merging occurring successfully at rate approaching 90% even with 450bp inserts.
600 cycle cassette? I don't know, is anyone getting reasonable 300 base reads with those?
--
Phillip
|