Sequencing beads outside recommended enrichment %

[rosaberg]

Hi, I am quite new to sequencing, but have been doing it the past 5 months with rather good results. I am wondering if anyone have sequenced beads significantly outside of Roches recommended range, i.e. 5-15%? I have lots of beads at 30% and would like to try...Is it likely not to work? What can happen? Would be grateful for any reply! Regards Eva

[Anthony.287]

I've sequenced beads with as high as 60% enrichment, but nothing over 20% has ever worked well here. I know of some users that won't sequence anything more than 10%. You'll probably end up with a high percentage of mixed and short quality reads, but should get some useable sequence as well, if you do end up sequencing those.
Good luck!

[rosaberg]

I've sequenced beads with as high as 60% enrichment, but nothing over 20% has ever worked well here. I know of some users that won't sequence anything more than 10%. You'll probably end up with a high percentage of mixed and short quality reads, but should get some useable sequence as well, if you do end up sequencing those.
Good luck!

Thanks Anthony, best to leave them then...Regards/Eva

[MissDNA]

The max we ever used was like 27% or so because we are in a hurry. The results were not bad and consistent with what we had obtained before for the same library with lower enrichment.

Other than that as a rule, we don´t sequence stuff above 20%

We do sequence stuff with more than 10% all the time, and we don´t notice bad results associated to that.

[tokikake]

we sequenced beads with an enrichment of ~30% (related to the recovered beads). When this value is reliable (and not a false good one) the sequencing results were good.

[genomeseeker]

Agreed - we typically use a 20% cutoff; the higher the enrichment rate the increased chance of mixed reads.

[ChristianBourne]

On a recent project of 16 Flx+ runs I found an enrichment of ~25% gave best results, but for a XLR70 run I'd stick to between 5-20%, although to be honest with you with all the difficulties we're experiencing I'd go for as low as possible...

[Aniki]

We have run as high as 24% and it was not bad. Recently though, it is hard to gage exactly what to load and what not to load. I have either not enough beads to load or I have enough to load and the raw well numbers are too low.

[mohdsos]

The maximum good enrichment I did was 33% and it was good.

[ChristianBourne]

You also have to consider that Roche are currently saying there may be an issue with 'unbound' DNA capture beads being carried through the enrichment and giving false enrichment values, this may also contribute to low numbers on the 454.

[bradfish]

You also have to consider that Roche are currently saying there may be an issue with 'unbound' DNA capture beads being carried through the enrichment and giving false enrichment values, this may also contribute to low numbers on the 454.

Can you clarify? By unbound, do you mean null beads, as in no DNA on them?

Also, how did you hear this? Roche has not come out and said this publicly, as far as I can tell.

[Aniki]

Can you clarify? By unbound, do you mean null beads, as in no DNA on them?

Also, how did you hear this? Roche has not come out and said this publicly, as far as I can tell.

You have to pay attention to the washing/melting steps for Enrichment. The White DNA beads they cling to the brown enrichment beads. This is not normal and should not be happening. When this is the case, you will *enrich* a set amount of null beads. This can cause an abnormally low number of raw wells.

[ChristianBourne]

Hi bradfish,
You betcha I can clarify baby! ;-)
Yes I mean 'null' beads. Our FSA told me personally. The number of washes in the protocol when enriching on the MPC may be unrealistic. But you know as well as I do that each sample can behave differently, and that's even before you factor in the differences in Lot numbers of kits!
It's yet another case of Roche not putting out a release of some kind informing all it's customers there may be an issue. I have spoken to fellow users who are not aware of any issues relating to inconsistencies with kits over the past few years or so. We all torture ourselves when we do everything correctly and our samples pass all the QC steps (including the new adaptor QC and post emPCR QC), when in fact it could be nothing we're doing.
Yes I know samples misbehave and don't work for a myriad of reasons, but cooperation from the company must be more forthcoming.
For example, I have had a recent order for 16 Flx+ runs. We have two Flx+ 454s. From four emulsion cups I enriched enough beads for four separate runs. I aloquotted my enriched beads out so all eight halves across my four runs contained exactly (sic) the same beads but I used two separate Lot numbers of reagents.
I saw a ~10% difference between machines and ~20% difference between Lot numbers.
Go figure. So factor that in with possible DNA capture bead manufacturing issues etc, and it adds up to a whole world of stress for us.
One thing you must do is check the wind direction and that you have your 'lucky' socks on when setting up a run. Now that's NGS science for you :-)

[Anthony.287]

@ChristianBourne - I wish I could "like" your comment.

[ChristianBourne]

Viva la revolution my friend!
I've got three years of stress, pain and misery built up thanks to Roche and their blooming 454. I used to be soooo happy running my GAiixs. Or am I just imagining that? The HiSeq and MiSeq I have work well enough, and there's no emulsion pcr
Maybe we should admit defeat and turn all the Flx+s back to titaniums.
Everyone is ooohing and aaaaahing over the 2x250 MiSeq, but come on it wouldn't touch a full 454 run for throughput. But the cost....