Tweet
Hi all,
I have a low complexity RNA sample (think a purified pool of a single gene) that I'm trying to prepare for sequencing using the TruSeq kit, and I'm ending up with large (typically >1kb) fragments, rather than the usual 300-400 bp fragments in a typical TruSeq prep.
I've traced the troublesome step to the RT, where the fragmented RNA is a distribution of sizes right around 200bp, but the cDNA is a much wilder distribution, with a strong peak around 900bp. A mixed pool of total RNA run exactly in parallel behaves normally (top rows in the attached bioanalyzer traces), giving size distributions consistent with normal libraries.
Any thoughts on what's going on here, and more importantly how to prevent it? The documentation I've found says "the tendency for SuperScriptTM II to switch templates is minimal since it is an RNase H minus enzyme", but I'm worried that with this super-low complexity sample, the fragments are somehow scaffolding each other to form a full-length version of the gene (I haven't yet determined whether this happens at the first or second strand synthesis yet), whereas in a more normal RNA sample, the odds that the correct fragments would link up are much smaller.
Thanks!
Peter Combs
I have a low complexity RNA sample (think a purified pool of a single gene) that I'm trying to prepare for sequencing using the TruSeq kit, and I'm ending up with large (typically >1kb) fragments, rather than the usual 300-400 bp fragments in a typical TruSeq prep.
I've traced the troublesome step to the RT, where the fragmented RNA is a distribution of sizes right around 200bp, but the cDNA is a much wilder distribution, with a strong peak around 900bp. A mixed pool of total RNA run exactly in parallel behaves normally (top rows in the attached bioanalyzer traces), giving size distributions consistent with normal libraries.
Any thoughts on what's going on here, and more importantly how to prevent it? The documentation I've found says "the tendency for SuperScriptTM II to switch templates is minimal since it is an RNase H minus enzyme", but I'm worried that with this super-low complexity sample, the fragments are somehow scaffolding each other to form a full-length version of the gene (I haven't yet determined whether this happens at the first or second strand synthesis yet), whereas in a more normal RNA sample, the odds that the correct fragments would link up are much smaller.
Thanks!
Peter Combs