[annaoc]

Hi everyone. With the ScriptSeq mRNA-Seq Library Preparation Kit, would you recommend the MinElute or gel electrophoresis for library purification?
Has anyone had experience of primer dimers after the PCR of 10 cycles? Cutting the band via size selection from a gel would perhaps then be the better choice. However, we do realise that by using the MinElute we can avoid exposing our sample to UV and Ethidium bromide and so we will probably obtain a much purer sample from MinElute as well as it being much a much easier protocol. Any comments on this would be greatly welcomed!
Also, will our library still include the tRNA and small RNAs? Our bioanalyser run of the rRNA depleted RNA has shown that we have quite a lot of tRNA and we were wondering if they will be in the library then we can avoid them by size selection on the gel?
Many thanks for your help!

[epibio]

We recommend using 1.2X Ampure XP beads, if you want to get rid of the small amplicons. MinElute is not as good at doing this; you can expect ~10% adaptor-related reads with MinElute as compared to <3% with Ampure.

[jagruh]

epibio: Is ScriptSeq mRNA-seq library prep kit (for Illumina) strand specific? From the mechanism, it seems strand specific, but in our sequenced result there are still enormous hits to the other strand. We wonder whether these anti-transcripts are biological (in bacteria) or artifacts.
Also, The distribution of hits within one gene is very uneven. It seems the PCR artifacts and/or non-random fragmentation inference a lot. Is there any available normalization parameters or tools available for this kit? And is there any published data prepared by this kit available in SRA database?
Thank you!

[arabidopsis]

Dear all,

I am analyzing the result of RNAseq on Hela cells. Libraries were created by Epicentre ScriptSeq kit, and they have around 7.000.000 reads each. I mapped the reads with TopHat to both hg18 and hg19. Most of transcripts look OK, but weird thing happens with RNA genes-snRNA, snoRNAs etc. Almost all reads cover reference RNAs starting from nucleotide +5 to +10, not from the annotated 5' end, even if the total coverage is very high. Out of more than 2000 reads mapped to U1 snRNA only 2 or 3 have their 5' end identical to the gene. Have anyone experienced this? Is there some technical issue with the library construction or sequencing, or a mapping artifact?
Best,

Aleks

[epibio]

@arabidopsis: The ScriptSeq method is designed for mRNA-Seq, and is probably not the best fit for small RNAs. It may be partly a mapping problem as well; you could try allowing for more mismatches.

@jagruh: Sorry I missed your post--for others who are interested, the ScriptSeq method is directional; however, we don't have data published to SRA as yet.

[upendra_35]

Dear all,

I am analyzing the result of RNAseq on Hela cells. Libraries were created by Epicentre ScriptSeq kit, and they have around 7.000.000 reads each. I mapped the reads with TopHat to both hg18 and hg19. Most of transcripts look OK, but weird thing happens with RNA genes-snRNA, snoRNAs etc. Almost all reads cover reference RNAs starting from nucleotide +5 to +10, not from the annotated 5' end, even if the total coverage is very high. Out of more than 2000 reads mapped to U1 snRNA only 2 or 3 have their 5' end identical to the gene. Have anyone experienced this? Is there some technical issue with the library construction or sequencing, or a mapping artifact?
Best,

Aleks

Hi could you please provide me your parameters for mapping strand-specific reads using Tophat. I ask this because when i map my reads i can only see 0.5% reads mapped. Thanks in advance

[mmark27]

Dear all,

I am analyzing the result of RNAseq on Hela cells. Libraries were created by Epicentre ScriptSeq kit, and they have around 7.000.000 reads each. I mapped the reads with TopHat to both hg18 and hg19. Most of transcripts look OK, but weird thing happens with RNA genes-snRNA, snoRNAs etc. Almost all reads cover reference RNAs starting from nucleotide +5 to +10, not from the annotated 5' end, even if the total coverage is very high. Out of more than 2000 reads mapped to U1 snRNA only 2 or 3 have their 5' end identical to the gene. Have anyone experienced this? Is there some technical issue with the library construction or sequencing, or a mapping artifact?
Best,

Aleks

I know this thread is wicked old, but we're using the Scriptseq V2 kit and we see the same thing on our data when loaded into the Genome Browser. Everything is shifted slightly. Epibio didn't have an explanation of why.

What are people using during TopHat for library type? fr-firststrand, fr-secondstrand????

[MPPFig]

libraries with Chloroplast rRNA contamination-Epicentre Ribo-Zero Kit

I used the ScriptSeq(TM) mRNA-Seq Library Preparation Kit v.2 to construct some Brachypodium libraries. mRNA was obtained using the Ribo-Zero Kit (Plant Leaf), it did not work so well, it left a 1300 nt fragment that appears to be chloroplast rRNA. Based on the Agilent High Sensitivity DNA assay my libraries have "spiky" look. The patterns might come from the fragmentation of that abundant 1300 nt seq. I am worried that I will end up wasting a lot of the reads by sequencing chloroplast rRNA. I am attaching the bioanalyzer images. Has someone experience this?

[diagen]

I get the same spiky pattern on my RNA samples, that also come after RiboZero!
I suspect something wrong with the RiboZero step, because on RNAseq prep on total mRNA have no such pattern.
Importantly, the yield of Ribozero is only 5% of total RNA, leaving 95% to rRNA. Nevertheless, textbook studies estimate the rRNA cellular content about 50%!

I wonder, where is the rest of my precious sample? Stayed on Ribozero beads?
What kinds of other biases the RZ kit does???

[MPPFig]

Diagen,

could you post some images after the RiboZero step, and bioanalyzer traces of the spiky library prep. What organism are you working on? Epicentre thinks that my 1300 nt fragment that you see after RiboZero is some rRNA that is not present on they cocktail. I am working with them trying to figure this out.

I am taking the rest of my total RNA (my so... precious sample) and try to clean with the Ambion MicroPoly(A)Purist™ Kit but the amount of total RNA that this kit requires is a concern, what if I lose everything?

[diagen]

Here are 2 typical profiles.

[MPPFig]

diagen,
I would talk to the epicentre tech support , one thing they did for me was to BLAST my rRNA sequences against their pool of RiboZero probes, that way they could assess how well their kit would work for your samples.

[MPPFig]

diagen,
I would talk to the epicentre tech support , one thing they did for me was to BLAST my rRNA sequences against their pool of RiboZero probes, that way they could assess how well their kit would work for your samples.

[ruwehann]

chloroplast 23S rRNA

Chloroplast 23S rRNA which is properly inserted into the ribosome gets cleaved at two position giving rise to fragments of typically 1.3, 1.0 and 0.5knt. For complete rRNA depletion one would need more than one probe for chloroplast 23S rRNA.

[diagen]

My samples are from human cell line. Furthermore, the first step in ScripSeq is non-specific RNA degradation by heat and divalent cations. So in principle no discrete bands are expected, but they are there in samples after RiboZero.