Amplicon sequencing

[henry.wood]

Another quick related question, has anyone tried or does anyone know about anyone who has tried the following?

1. PCR with specific primers.
2. Shearing of PCR products.
3. Ligation of adapters (optionally indexing).
4. Sequencing.

instead of

1. PCR with adapted primers.
2. Shearing.
3. Sequencing.

I am far more interested on the first method. Any comments? Any ideas?

thanks in advance!

Hello. Some colleagues of mine routinely do the former. http://www.ncbi.nlm.nih.gov/pubmed/20127978 is what they were doing a little while ago, although they have probably tweaked things a bit since then.

[bbeitzel]

We have a current project where we are generating two different amplicons from the same sample (~1.2 and ~1.4 kb), mixing those, then shearing by Covaris. We then run the sheared DNA through the TruSeq protocol to generate libraries. This works pretty well with these amplicons.

We have another project where we are generating smaller amplicons (both ~400bp). We have been able to get to this to work as well, but we had to tweak the Covaris conditions to get the small amplicons to fragment efficiently.

We have also played around with enzymatic fragmentation (Fragmentase from NEB and Shearase from Zymo), and I think with a little effort that would work as well. I did notice that on our small amplicons the Fragmentase gave a better fragment distribution than Shearase. Fragmentase gave a nice smear <400 bp while Shearase showed definite banding. Shearase is a very high frequency cutter (I think it is a 1.5-cutter,) but our small amplicons may have limited sequence heterogeneity resulting in the visible bands.

[CPCantalapiedra]

thank you very much for your answers!!

I'm still wondering if some has tried these methods with 454 sequencing,
so I'm forwarding this topic to the 454 forum.

[CPCantalapiedra]

Amplicon sequencing

A quick question, has anyone tried or does anyone know about anyone who has tried the following with 454?

1. PCR with specific primers.
2. Shearing of PCR products.
3. Ligation of adapters (optionally indexing).
4. Sequencing.

instead of

1. PCR with adapted primers.
2. Sequencing.

I am far more interested on the first method. Any comments? Any ideas?

thanks in advance!

[nilshomer]

Please do not make duplicate posts: http://seqanswers.com/forums/showthr...2223#post52223

[ECO]

Have you read this document?

https://shell.cgrb.oregonstate.edu/sites/default/files/Files/Docs/Roche/misc/454SequencingSystem_GuidelinesforAmpliconExperimentalDesign_May2011.pdf

...seems like you're describing "Design 3: Ligated Adapters".

Your second option doesn't make sense to me, unless it's long range PCR and you don't care about anything but the ends...if you've adapted your primers, why shear?

[ECO]

Merging in all responses from your thread-hijack in the ILMN forum...

[CPCantalapiedra]

Yes, you are right, no shearing at all on second option.
Thanks for that document, I'll take a look carefully

I don't understand why were considered as duplicated posts because one was in the 454 subforum and the other one in the Illumina, but... as you wish, of course.

[ajthomas]

The first method would really only be useful and make sense if you're doing long range PCR--something over ~1kb. Otherwise, just stick the adapter sequences on your primers, do the PCR, and sequence. It's much simpler, cheaper, and more reliable to put the adapter sequences in your primers.

[CPCantalapiedra]

The first method would really only be useful and make sense if you're doing long range PCR--something over ~1kb. Otherwise, just stick the adapter sequences on your primers, do the PCR, and sequence. It's much simpler, cheaper, and more reliable to put the adapter sequences in your primers.

Thank you for your advice ajthomas, but as you have stated, I'm looking for processing long sequences (maybe 2 kb the longest), and, if possible, I would like to sequence several amplicons (around 4*24 = 96) for as much as samples as I could (say a maximum of 192 samples). That's a lot of primers and PCRs doing it by the basic procedure.

I have taken a look at the alternative designs, and I really don't understand why (as it seems) Lib-A adapters can't be ligated for the LR-PCR design (number 4 in the document) instead of the Lib-L adapters. If anyone could explain this I would be very pleased.
thank you

[ajthomas]

If I'm reading that right, you have 96 amplicons, right? Or is that 4 amplicons with 24 MIDs? Either way, that is a lot of primers if you use the basic method. Since your amplicons are large, then fragmentation and ligation is probably the best method. If you can make the amplicons smaller, such that you don't need to fragment them, there is a way to attach large numbers of MIDs without buying separate primer sets for each one. Have a look at Fluidigm's Access Array system. I've been using it for a few months now, and it works pretty well. The way that it attaches MIDs is by using 4 primers, rather than two. One set of primers has the gene-specific sequence and an adapter sequence (they call it CS1 or CS2). Then, a second set of primers binds to CS1 or CS2 and adds the MIDs and 454 adapter sequences. That way, you don't need primers with each combination of gene-specific sequence and MID.

That approach only works if you can make the amplicons shorter. If you can't, or if you prefer to do the long PCR and ligate adapters on, that should certainly work. That's a lot of libraries if you have 192 samples, and a lot of expense. My guess is that it will be cheaper and less work to integrate the adapters in the PCR (assuming it's possible to shorten the amplicons, of course), rather than ligate them on, but that's something for you to figure out and decide on your own.

As for ligating on Lib-A adapters, I can't see any reason why you couldn't. If you were to do that, however, you would have to make the adapters yourself since 454 doesn't sell them. Also, there wouldn't be much point. If you're ligating on adapters, you will be sequencing in both directions anyway, so using the Lib-A adapters wouldn't give you any benefit.

[CPCantalapiedra]

thank you very much for your advice ajthomas

If you can make the amplicons smaller, such that you don't need to fragment them, there is a way to attach large numbers of MIDs without buying separate primer sets for each one. Have a look at Fluidigm's Access Array system. I've been using it for a few months now, and it works pretty well. The way that it attaches MIDs is by using 4 primers, rather than two. One set of primers has the gene-specific sequence and an adapter sequence (they call it CS1 or CS2). Then, a second set of primers binds to CS1 or CS2 and adds the MIDs and 454 adapter sequences. That way, you don't need primers with each combination of gene-specific sequence and MID.

Thanks very much for the info. I'll take a look. I was wondering if this can be done mixing the repaired fragments of all the amplicons of each sample and then ligating the RL adapters with the MIDs.

That approach only works if you can make the amplicons shorter. If you can't, or if you prefer to do the long PCR and ligate adapters on, that should certainly work. That's a lot of libraries if you have 192 samples, and a lot of expense. My guess is that it will be cheaper and less work to integrate the adapters in the PCR (assuming it's possible to shorten the amplicons, of course), rather than ligate them on, but that's something for you to figure out and decide on your own.

umm I'm not sure if would be cheaper and less work to do one PCR with specific primers and MID tagged adapters for so many samples... say 96 PCR (amplicons) * number of samples, or 96 PCR + number of samples PCR (if I could pool the amplicons of each sample to do the tagging). And ligation would give reduced bias, as far as I know (note that I'm novice, of course).

As for ligating on Lib-A adapters, I can't see any reason why you couldn't. If you were to do that, however, you would have to make the adapters yourself since 454 doesn't sell them. Also, there wouldn't be much point. If you're ligating on adapters, you will be sequencing in both directions anyway, so using the Lib-A adapters wouldn't give you any benefit.

umm I am missing something here, because I thought Lib-L was just for unidirectional sequencing, but as you say in the document states that "Obtain reads from both strands from a single Adaptor / MID " as an advantage of the LR-PCR design. Then, what's the difference with using the one-way reads? It's a shame in the document there are missing figures... for beginners to figure out!

[CPCantalapiedra]

anyway it seems clear that the number of samples and amplicons is too big, so I consider this PCR based approach just in case of reducing the number of samples or amplicons... but thank you anyway for the advices, because maybe 96 amplicons with 8 pools is not so frightening

[ajthomas]

umm I'm not sure if would be cheaper and less work to do one PCR with specific primers and MID tagged adapters for so many samples... say 96 PCR (amplicons) * number of samples, or 96 PCR + number of samples PCR (if I could pool the amplicons of each sample to do the tagging). And ligation would give reduced bias, as far as I know (note that I'm novice, of course).

You certainly can, and in fact should, pool the amplicons before doing the tagging. Before doing so, though, you need to accurately quantify your PCR products and then use that information to pool them stoichiometrically. In fact, you would do that whether you tagged the amplicons post-PCR or integrated the adapters into the primers.

As far as ease and cost, it all depends on numbers. Each library you make by ligation will cost you $100-200, whereas the primers will cost you about $10 each. (Of course, those are US list prices; I don't know what it is in Spain.) With those primers, you can do as many PCRs as you like. If you have many amplicons, but only one or two samples, it will be more cost effective to just tag the amplicons after the PCR, but if you have only a few amplicons and many samples, it's cheaper to integrate the adapter sequences.

It gets a little more complicated when you add in MIDs. If you will be pooling many samples, necessitating the use of many MIDs, you will need a separate primer pair for each sample, and that may raise the cost considerably. However, if your amplicons are short enough, (say <400 or so with the current FLX, maybe <700 with the new FLX+) that most of your reads will be full length, you can use different MIDs on each end and use many fewer MIDs. In your case, if you have 96 samples you want to sequence in the same pool, you would need 192 primers (96 forward, 96 reverse) if you wanted to sequence from both ends and the amplicons were too long to sequence clear through. If they are short enough that you can use different MIDs on each end, you could do it with only 20 primers (10 each direction). My mixing the MIDs, you could encode 100 (10X10) different samples with those 10 MIDs on each end.

umm I am missing something here, because I thought Lib-L was just for unidirectional sequencing, but as you say in the document states that "Obtain reads from both strands from a single Adaptor / MID " as an advantage of the LR-PCR design. Then, what's the difference with using the one-way reads? It's a shame in the document there are missing figures... for beginners to figure out!

Lib-L is for unidirectional sequencing, but only when the sequences are integrated into the PCR primers. If they are ligated on to the ends of the fragments, will sequence in both directions with Lib-L. The reason is because if you integrate the adapter sequences into the primers, you will always be sequencing from the primer with the A adapter, and unless you make two sets of primers with the A and B switched, the A adapter and therefore your sequence, will always be on the same end. If the adapters are ligated on, some molecules will have the A adapter on one end and others will have the A adapter on the other end.

[CPCantalapiedra]

Thank you very much for your answers, I'm really impressed with your disposal

You certainly can, and in fact should, pool the amplicons before doing the tagging. Before doing so, though, you need to accurately quantify your PCR products and then use that information to pool them stoichiometrically. In fact, you would do that whether you tagged the amplicons post-PCR or integrated the adapters into the primers.

If they are short enough that you can use different MIDs on each end, you could do it with only 20 primers (10 each direction). My mixing the MIDs, you could encode 100 (10X10) different samples with those 10 MIDs on each end.

That's a great idea, but I guess would be 10!/(8!*2!) + 10 = 55, not 10*10??

Lets say I have 64 samples. With the amplicons, say I have 2 options:

1) Create >1kb amplicons. Say 24 loci * 4 amplicons = 96.
2) Create ~300 b amplicons. Say 8 loci * 12 amplicons = 96.

needing 96 primers in both cases for the initial PCR.

1) number of PCRs = 96*64. (or maybe could be pooled in this step?)
Then shearing and repair.
Then pool (so I got 64 samples).
Ligation RL adapters to each sample using MIDs. So, this is the step you say would cost 100-200 $ each sample? I mean: 6400-12800$ in total?

2) number of PCRs = 96*64 (or maybe could be pooled in this step?),
using primers with the universal tail.
Then PCR step 2, to include the MIDs and adapters = 64 PCRs,
with 12 MIDs (this would be 12!/(2!*10!) + 12=78 ¿?)
being 12 * 2 = 24 primers.
Even I could do both PCRs in just one step... although maybe to messy

It seems obvious that 2) is the best option, but I have lost a lot of loci, and incresing it
would mean to increase the number of PCRs being maybe unmanageable (24 * 12 = 288 amplicons --> 288 * 64 PCRs?!?!)

I'm going to take a serious look at targeted sequencing.

Lib-L is for unidirectional sequencing, but only when the sequences are integrated into the PCR primers. If they are ligated on to the ends of the fragments, will sequence in both directions with Lib-L. The reason is because if you integrate the adapter sequences into the primers, you will always be sequencing from the primer with the A adapter, and unless you make two sets of primers with the A and B switched, the A adapter and therefore your sequence, will always be on the same end. If the adapters are ligated on, some molecules will have the A adapter on one end and others will have the A adapter on the other end.

That actually makes sense, when I have time I should do that PCR on paper to learn it well
but opens for a new question, why not using just adapter A on both primers? because of the single stranded steps?

thanks again for your help, it's invaluable