SOLiD v3/v4 vs. 5500

[gendxdoc]

Has anyone compared RNA-seq data generated on a SOLiD v3/v4 with a 5500 -- for the same sample ?

We've recently taken the same total RNA sample, constructed a 5500 library and performed a 50bp run on it. A comparison of the data (v4 vs. 5500), looks considerably different. On it's own -- the 5500 data looks great ! But the correlation between the two data sets is very poor. A handful of qPCR expression assays -- tends to confirm the 5500 data.

Does anyone have any experience with this ? Several changes have taken place with the 5500 upgrade -- new library construction protocol ("T/A" tailed ligation, bead clean-ups, etc.), updated 5500 chemistry, new instrument,etc.
One would expect higher quality data -- and more of it. What's disconcerting is how different it looks from the v4 data.

Any ideas would be greatly appreciated !
Thanks,
Michael

[pmiguel]

Michael,
Eh?

ABI uses the Ambion RNA library kit. It directly ligates RNA to double-stranded adapters with 5' or 3' five random nucleotide overhangs. It would not be compatible with T/A cloning.

Are you sure you (or your sequence provider) did not switch to "create cDNA then fragment and treat as a DNA library" to save money on library construction? Obviously this could result in a different sequence profile.

By the way, ABI really, really needs to drop their library kit prices to have any hope of competing with Illumina since the new TruSeq kits were released. The Ambion "SOLiD Total RNA Seq Kit" is directional, while the TruSeq RNA kit is not. But we are talking >$400 (if you count ribosomal depletion -- included in TruSeq, but not in Ambion kit) vs ~$70 for TruSeq in reagents so there is a big problem for ABI here if they don't respond soon.

--
Phillip

[gendxdoc]

Sorry --- didn't supply you with enough information.

What I'm doing is using a single cell (or few cell) protocol for generating cDNA from total RNA -- been doing this for quite a while, works very well.

The cDNA is then sheared and a fragment library prepared from this material. The previous protocol (v3/v4) used a blunt ligation to add adapters and columns for each purification step. The new 5500 fragment library protocol uses T-tailed adapters, with bead-based purification. Much more efficient.

Michael