[Tirtha]

Density of oligos in the flow cell V3

Hi Friends,

I am new to illumina hi seq and i am in the process of understanding the technology for its application towards exploring small RNA diversity. Please educate me on the following issues---
1. No. of oligos present in the illumina flow cell V3 per lane.
2. How many reads are considered ideal for diversity/expression studies
3.What are the recommended procedures/ precautions needed to reduce redundancy and enhance normalization of data across the various stages of sample preparation and sequencing protocol.
4. And yes, if i am using enriched small RNA , then how much of it desirable for such studies(available literature only talks in terms of total RNA)

Thank you very much!!

[ScottC]

I'm currently analysing cufflinks output from around 10 millions reads to up to 240 millions so I guess I'll see first hand how many reads we need in our system.

Hi pettervikman,

Any updates?

Scott.

[pmiguel]

Sure.
BTW, does anyone have a citation to a paper that shows that DSN normalization does not bias strongly the transcripts expressed at lower levels than the ribosomal? I find:

http://nar.oxfordjournals.org/content/39/20/e140.full

But I thought I remembered seeing something more exhaustive?

--
Phillip

[pettervikman]

Thanks a lot for all the answers.

The reason I went for the poly A tail was that from a statistical stand point a certain fraction of the reads will be from the 3' end of the transcript and these could (should ?) have a few As ot Ts at the end since the fragmentation woun't be exactly between the tail and the transcript. So then there could be an increase in A/T in the beginning of the reads. That said there should also be a similar increase in A/T at the end of fragments since we are using an unstranded setup but this I overlooked.

So thanks for informing me about this.

Anyhow, I'm currently analysing cufflinks output from around 10 millions reads to up to 240 millions so I guess I'll see first hand how many reads we need in our system.

/Petter

[cedance]

The link I pointed to, basically tells the same story. Phillip explained it crisply and well.

[pmiguel]

No, I mean using one of the cDNA normalization via double stranded nuclease(DSN) methods. You create double stranded cDNA, denature the strands, allow re-annealing to occur for an interval long enough for highly expressed transcript strands to find one another, but not those expressed at a lower level. Add DSN. Purify. Re-synthesize 2nd strand, continue as normal.

There was a method paper published showing the validity of this method for RNAseq. Can't seem to find it now, though.

--
Phillip

[cedance]

I guess the normalization pmiguel talks about is experimental. One of my colleagues explained something about library normalization using enzymes. It seems that the enzymes will act on cDNA-mRNA complex degrading them. Its probably explained better here, not sure though. Since the highly expressed genes are abundant, they'll be depleted in larger quantity. (I am not a biologist. So pardon my terminology).

@pmiguel, about the fact that the bias is due to random hexamer priming, this is what I have heard of as well. Thanks for pointing that out.

@petter, I am not sure about the poly-A story. 1) I don't think there is a necessity for excess A/T at the 3' end. 2) Even so, you are then fragmenting and amplifying. So, why would they still be selectively amplified?

[pettervikman]

pmiquel: I've thought about depleting them through the same technical pipeline as is used for rRNA depletion for example. This woun't affect the actual sequencing though, I would not get more reads for the lower expressed genes. But by a physical depletion where I actually removed these transcripts I'd get more reads and maybe more transcripts sequenced.

But I guess that your talking about the data depletion rather then the transript depletion?

[pmiguel]

I'm also curious whether it's much dependent on the highly expressed genes that are in the sample since they "steal" a lot of the data being produced. I know that it's possible to select the genes that one is interested in but have any one tried to remove the genes that is uninteresting/highly expressed to increase the coverage of the other genes? This would allow for a higher coverage even of genes that you don't know exist in comparison to the positive selection when you only find what you expected to find.'

You can deplete the highly expressed genes globally using normalization. My naive presumption would be that this would distort the relative numbers of all the transcripts. But recent studies have shown this not to be the case.

--
Phillip

[pettervikman]

Thanks pmiquel. Didn't know that. But I've heard that it's much more common in rna-seq experiments in comparison to dna seq, hence the poly a tail story. But your saying that it's only dependent on the 2nd strand syntesis?

[pmiguel]

Thanks again. I am sorry I don't/haven't used cufflinks, yet.
1 more question!!: why is poly-A pulldown responsible for initial increase in A/T?

It isn't. The non-random base distribution in the first 10 bases is attributed to hexamer-primed 2nd strand synthesis. (The hexamers do not prime perfectly randomly.)

--
Phillip

[pettervikman]

Thanks! They got sequenced here on "my" hiseq. We have a hiseq here on CRC in Malmö, and where part of Lund University/LUDC (Lund University Diabets Center).

The pulldown uses a poly T tail and this will bind somewhere in the poly A tail (just to be super clear). Hopefully close to the 3' end of the CDS/3'non coding. But if it binds further down there will be a few As or Ts sequenced before the actual sequencing, hence the slight increase of A/T.

[kopi-o]

Petter, those data look super. Did you get them sequenced in Uppsala?

[cedance]

Thanks again. I am sorry I don't/haven't used cufflinks, yet.
1 more question!!: why is poly-A pulldown responsible for initial increase in A/T?

[pettervikman]

Here are per base content and duplication levels. Since we've used the poly A tail pulldown I'm not suprised of the increase in A/T initially. The duplication levels are much higher then I'd accept for a genomic project but since there's much less diversity from the transcriptome I'm fine with this. Consider that there are hard end points that really cant be changed (5' and 3' ends of transcripts) and between maybe 10-15 k transcripts to start with.

An other question though. After cufflinks using RABT (-g) the transcripts creation looks a lot nicer. That said does anyone know why some transcripts are labelled OK despite the fact that their FPKM_low is 0? I'm also wondering about transcripts labelled as FAIL that have the positive numbers in coverage, fpkm, fpkm_high.

To sum it up, why are there transcripts with positive numbers in coverage, fpkm, fpkm_high and 0 in fpkm_low sometime OK, LOWDATA or FAIL?