[pmiguel]

Do you use different standards to quantitate based on the GC content of the genome or based on the size of the genome?

No. Most of the time these figures are "ballpark" anyway. 50% high or low will make little difference.

Phenol or RNA, though, can easily contribute >90% of the UV absorbance of a genomic DNA sample. 10-fold low on the amount of input DNA does make a difference.

--
Phillip

[LMcSeq]

Hi all,
Quick question: I'm finding the my average frag length on the high sensitivity chip at the end of the rapid protocol is on the larger side. My sequencer results have had low average fragment lengths. I'm wondering if the larger frags aren't being amplified because they're too big for the microreactors, causing preferential amplification of smaller frags. I'm using the Covaris settings recommended for 500bp average. Anyone else having any similar issue?

[pmiguel]

Hi all,
Quick question: I'm finding the my average frag length on the high sensitivity chip at the end of the rapid protocol is on the larger side. My sequencer results have had low average fragment lengths. I'm wondering if the larger frags aren't being amplified because they're too big for the microreactors, causing preferential amplification of smaller frags. I'm using the Covaris settings recommended for 500bp average. Anyone else having any similar issue?

It could be. Typical PCR reactions yield vastly more product for short templates than for longer ones.

But it could also be that your read lengths are short for some other reason. If your templates are actually short you should have a high number in your "Short Primer" metric. If not, then something else is likely the source of your short reads.

I will mention one: polyA, of course, is the bane of 454 runs...

Even with random primed cDNA libraries, it is possible to run into polyA problems. It should be less, but polyA+ RNA isolation will enrich for polyA. If you start with heavily degraded RNA and pull out polyA from that, you may get a high percentage of your cDNA being polyA or polyT--even if you used random primed reverse transcription.

--
Phillip

[LMcSeq]

Hi Phillip,
Thanks for your reply. We've had our 454 FAS out and he said that its likely a problem in library prep or emPCR based on the output from the support tool (which customers can't view themselves). He said the short primer is in spec.

We are shearing whole genomic microbial DNA, not cDNAs. I think that was another contributor to this thread.

[pmiguel]

Hi Phillip,
Thanks for your reply. We've had our 454 FAS out and he said that its likely a problem in library prep or emPCR based on the output from the support tool (which customers can't view themselves). He said the short primer is in spec.

We are shearing whole genomic microbial DNA, not cDNAs. I think that was another contributor to this thread.

Okay, but if short primer is in spec, that means you do not have a large number of reads that traverse the insert completely. So preferential amplification of your small templates is not the issue.

It could be that you have too many templates longer than the maximum optimum amplification length of the emPCR kit (~1 kb, I think). Those would tend to produce lightly templated beads that might give you short reads.

--
Phillip

[LMcSeq]

That's what I suspect is the problem, but I'm not sure how to effectively troubleshoot it with the new rapid protocol since the chip isn't run until the end of all of the steps are complete. I don't want to have to make libraries over and over...
I'm not sure if I should do a calibration with my XP beads and check the avg. frag length with different Covaris shearing times (don't know if it translates to the new procedure with the sizing solution and standard XP volume).
Any suggestions?

[pmiguel]

That's what I suspect is the problem, but I'm not sure how to effectively troubleshoot it with the new rapid protocol since the chip isn't run until the end of all of the steps are complete. I don't want to have to make libraries over and over...
I'm not sure if I should do a calibration with my XP beads and check the avg. frag length with different Covaris shearing times (don't know if it translates to the new procedure with the sizing solution and standard XP volume).
Any suggestions?

We would just size fractionate with an agarose gel if we got a library too much outside the size range. I realize that is not optimum for large numbers of samples, though.

Also, why not use nebulization instead of sonication? I believe nebulized DNA would have a higher fraction of ligatable ends after end-repair than sonicated DNA.

--
Phillip

[kmcarr]

Okay, but if short primer is in spec, that means you do not have a large number of reads that traverse the insert completely. So preferential amplification of your small templates is not the issue.

Phillip, reads will only be reported as short primer if they are rejected because they are too short (≤ 84 flows, ~50nt) after trimming off the B-adapter sequence. If an insert is 250nt long the raw read will likely traverse into the adapter which will be trimmed from the final read output. This will not be included in the shortPrimer filter count, that only reports reads which are rejected, not simply trimmed.

[seqAll]

Hi all,
Quick question: I'm finding the my average frag length on the high sensitivity chip at the end of the rapid protocol is on the larger side. My sequencer results have had low average fragment lengths. I'm wondering if the larger frags aren't being amplified because they're too big for the microreactors, causing preferential amplification of smaller frags. I'm using the Covaris settings recommended for 500bp average. Anyone else having any similar issue?

Hi LMcSeq,

Do you have leftover enrichment beads? If so, you can run normal PCR with a few of these beads as template, using the amplification primers, for a few cycles (with the same cylcling condition as emPCR, and maybe one or two cycles, so not to skew the size distribution). Then run on a gel to check the size distribution.

Next time, you can save the supernatant after melting (the supernatant are single-stranded templates released off the beads), and run them on a gel as a control step before you proceed to sequencing.

It would be interesting to see if it is the sequencing part (or maybe the algorithm) that can not take such long fragments.

[pmiguel]

Phillip, reads will only be reported as short primer if they are rejected because they are too short (≤ 84 flows, ~50nt) after trimming off the B-adapter sequence. If an insert is 250nt long the raw read will likely traverse into the adapter which will be trimmed from the final read output. This will not be included in the shortPrimer filter count, that only reports reads which are rejected, not simply trimmed.

Wow, thanks for that information. Seems absolutely crazy for Roche not to provide a metric or graph that shows the length of library molecules for which a B-adapter sequence was located. With that metric you can immediately see many library issues. Without it you can waste your time looking for instrument problems.

I have been trying to convince Applied Biosystems of the same thing.
--
Phillip

[kmcarr]

Wow, thanks for that information. Seems absolutely crazy for Roche not to provide a metric or graph that shows the length of library molecules for which a B-adapter sequence was located. With that metric you can immediately see many library issues. Without it you can waste your time looking for instrument problems.

Yes, it certainly would be nice. To check for B-adapter I output the untrimmed FASTA from the SFF file (sffinfo -s -n) and then search the reads against the B-adapter sequence with cross_match or fuzznuc (or your favorite sequence search tool).

[pmiguel]

Yes, it certainly would be nice. To check for B-adapter I output the untrimmed FASTA from the SFF file (sffinfo -s -n) and then search the reads against the B-adapter sequence with cross_match or fuzznuc (or your favorite sequence search tool).

Yes.
As it turns out cross_match and fuzznuc are my favorite sequence search tools.
Have you tried to see long (read-killing) homopolymer runs this way? Or do they not show up in even the untrimmed sequence?

--
Phillip

[bia]

Hi everybody,
does anyone experienced emPCR inhibition problems when using shotgun rapid libraries?
Thanks,
bia

[Cambridge454]

Rapid Library length problems

I am also having problems with the Rapid Library producing really long average fragment lengths >900bp. I use the recommended nebulization settings, which worked perfectly well for old version Shotgun libraries (600bp avg.). I have used the new vented caps and the rubber stopper/filter nebulizer set-up with very similar results. Should I increase the nitrogen pressure to say 2.5 bar instead of 2.1 bar?

[LMcSeq]

Cambridge454--I get long frag lengths with using the Covaris as well. If you adjust the nebulization, I would add more time rather than change the pressure. Try doing a time course study (that's what I did with the Covaris). Shear at the same pressure for different amts of time and assess on the Bioanalyzer. Unfortunately, I think the longer fragment lengths are just part of the rapid library process. I'm just hoping it doesn't cause preferential amplification of smaller frags since the larger ones may not fit into the microreactor bubbles during emPCR.