MiSeq Custom Sequencing Primer Tm?

[pmiguel]

Strangely Illumina does not want to part with this information. What they will tell you (I saw it in a FAQ or Tech document somewhere) is the following:

the MiSeq performs deblocking at 60 oC and incorporation at 65 oC (the HiSeq uses 55 oC for both those steps).

So, basically, "you know the Tm you use on the Hiseq/GAIIx sequencing primers that we refuse to reveal the sequence of, even though the information was freely available at some time in our history? Well you need to go about 10 oC higher than that."

Possibly this is because they have some extra chemistry in their MiSeq primers that increases the stability of the primer binding, rather than just adding some extra bases.

Anyway, I am sure many of you use custom primer on your MiSeq. What Tm do you shoot for? (And using what method of Tm calculation.)

--
Phillip

[mcnelson.phd]

Rob Knight's group worked with Illumina in designing and developing their 16S sequencing format, which uses custom primers for read 1/2 and index sequencing. Here's what I just calculated for the Tm of each:

Read 1: 65.36-66.93 (TATGGTAATTGTGTGCCAGCMGCCGCGGTAA)
Index: 60.71-64.75 (ATTAGAWACCCBDGTAGTCCGGCTGACTGACT)
Read 2: 60.71-64.85 (AGTCAGTCAGCCGGACTACHVGGGTWTCTAAT)
Ranges are due to ambiguous bases in the sequence.

We tried a modification of those primers to sequence a different 16S region on the HiSeq once, but the resulting data turned out very poorly. Not sure if that was our fault or the sequencing center since they were never that great at running things, which is why we got our MiSeq. We'll probably be trying new custom read/index primers once the software issues on amplicons gets sorted out.

[pmiguel]

Thanks mcnelson.phd,

I have a little more info from tech support:

the highest temperature the MiSeq uses during sequencing is 65C, and the highest temperature on the HiSeq is 55C. Based on that, I would design your custom primers to have a Tm above those temperatures, so that they will not melt off the clusters during the sequencing cycles.

Further, the Tm of our PE pre-TruSeq sequencing primers is 65.5C for the read 1 primer, and 70.1C for the read 2 primer, as calculated by IDT’s OligoAnalyzer. These primers are compatible with both our MiSeq and HiSeq systems. Since Tm calculation methods can vary, I would suggest using IDT’s OligoAnalyzer to compare the Tm of our PE pre-TruSeq sequencing primers to your custom designs for the highest accuracy.

PE Read 1 Sequencing Primer
5' ACACTCTTTCCCTACACGACGCTCTTCCGATCT
PE Read 2 Sequencing Primer
5' CGGTCTCGGCATTCCTGCTGAACCGCTCTTCCGATCT
Oligonucleotide sequences © 2007-2012 Illumina, Inc. All rights reserved.

Please note ECO that by including the copyright notice, I believe I am in compliance with the "limited permission" granted by Illumina in their "Illumina Customer Sequence Letter" document. Specifically, the "limited permission to distribute the sequence information outside your institution or to publish the sequence information in presentations, manuscripts, or publications authored by you, as long as it is accompanied by the following copyright notice". However, feel free to redact the sequences should you get any flack from Illumina.

Note that my earlier claim that Illumina refused to release the sequence of primers that were at one time published by them, is shown to be incorrect by the Illumina Customer Sequence Letter.

--
Phillip

[ZAAB]

We use a plethora of custom sequencing primers for the MiSEQ, and we shoot for 75-80 degrees Tm...I have not tested lower Tm primers, not worth the risk. We just pay for the longer oligo synthesis, which is still cheaper than a failed MiSEQ run.

[pmiguel]

Thanks ZAAB,
Exactly the sort of information I was looking for.

--
Phillip

[LSAD]

Hi ZAAB, have you used any custom sequencing primers that largely overlap with illumina p5/p7 adapters?

Originally Posted by ZAAB

We use a plethora of custom sequencing primers for the MiSEQ, and we shoot for 75-80 degrees Tm...I have not tested lower Tm primers, not worth the risk. We just pay for the longer oligo synthesis, which is still cheaper than a failed MiSEQ run.

[ZAAB]

If you're talking about the actual oligos on the lawn when you say the P5/P7 region, NO, we do not design the sequencing primers to anneal in that region...in our system there is no need for that length.
We've never deviated beyond re-building the traditional DNA sequence of the standard Illumina 5' adapter (rebuilt by PCR-installation of that sequence (tailing PCR)... and our PCR-derived constructs we're sequencing provide ~50nt 5'-to the region needing to be sequenced...so we typically use about 10-20nt of the illumina adapter sequence to extend primer annealing to achieve the correct Tm.
This leaves ~40nt of sequence before the 5'-most end of the construct, and 20nt or so between the 5' end of my sequencing primer and the region called P5/P7...the lawn-oligo-annealing-regions.

BUT there should be no issue extending the sequencing primer into that part of the adapter...

[charel]

I am working on the MiSeq.

I have designed custom sequencing primers to be in the ranges of the standard Illumina sequencing primers: length = 33bp, GC content = 51.5% and Tm = 65.8 ºC (from OligoAnalyzer). ZAAB, suggested that the Tm should be higher than that, in the range of ~75ºC and 80ºC, in order to avoid them falling off between at least initial cycles. Should I change the primers in making them longer but perserve their GC content?

A second related question I have is how much of a constraint hairpin formation or self dimers are on the design. My primer sequences contain an 8bp restriction enzyme site which can form the self-dimers and I am unsure whether that will be a problem or not.

Thank you for you help.

[pmiguel]

Originally Posted by charel

I am working on the MiSeq.

I have designed custom sequencing primers to be in the ranges of the standard Illumina sequencing primers: length = 33bp, GC content = 51.5% and Tm = 65.8 ºC (from OligoAnalyzer). ZAAB, suggested that the Tm should be higher than that, in the range of ~75ºC and 80ºC, in order to avoid them falling off between at least initial cycles. Should I change the primers in making them longer but perserve their GC content?

A second related question I have is how much of a constraint hairpin formation or self dimers are on the design. My primer sequences contain an 8bp restriction enzyme site which can form the self-dimers and I am unsure whether that will be a problem or not.

Thank you for you help.

I think you just go for the longest oligo possible that won't anneal to other places in your construct.

About the palindromic sequence -- that is kind of interesting. For Sanger sequencing I would expect terrible results for a sequencing primer that ended in a long palindrome. You get some percentage of the oligos annealing to each other and generating reaction products that migrate right on top of your normal products, obscuring the read. But for this solid state (templates attached to the flowcell) you are in better shape, no? All the primers that do not anneal would just get washed away prior to scanning. You might take a hit on intensity -- or not.

--
Phillip

[kkrishnan]

Hi, has anyone had experience with pooling custom sequencing primers? For instance if I am aiming to sequence different regions of the 16s, can I pool the sequencing primers. I wouldnt do this routinely but I am trying to validate different sets of primers and I'd prefer to use just the one cartridge at this stage. Since this is a trial run, I wont have many samples, so I can ensure unique index sequences for every primer set. FYI, the primers are modified from Caparaso et.al but targeting multiple regions.
Thanks, appreciate any suggestion/guidance.

[mcnelson.phd]

Originally Posted by kkrishnan

Hi, has anyone had experience with pooling custom sequencing primers? For instance if I am aiming to sequence different regions of the 16s, can I pool the sequencing primers. I wouldnt do this routinely but I am trying to validate different sets of primers and I'd prefer to use just the one cartridge at this stage. Since this is a trial run, I wont have many samples, so I can ensure unique index sequences for every primer set. FYI, the primers are modified from Caparaso et.al but targeting multiple regions.
Thanks, appreciate any suggestion/guidance.

The simple answer is to pool them in the same proportion as your different sets of amplicons, making sure that the concentration of the final pool corresponds with what's listed in Greg's paper.

The complex answer is that it depends on the length of each set of amplicons, such that shorter amplicons will cluster better and thus be more represented than larger amplicons. Primers are supplied in excess so that may not matter too much, but it was something that we thought of as being a possible problem when we tried this on a HiSeq run.

Lastly, you might want to be careful if you have primers that can potentially anneal internally to where the correct position is. We think we had this problem during the above mentioned HiSeq run and it ended up that one whole amplicon set was total crap. We didn't validate it on it's own, so it could have been just been a bad design, but I'm not sure if it's worth the risk. The new nano kits that will be coming out soon seem perfectly suited to testing out multiple ideas like this though.

[kkrishnan]

Thanks McNelson! Will take these points into consideration before I do my run. What are the new nano kits?

[kkrishnan]

Ah just found the update! Thanks.

[Andrew_Slatter]

Note that Fluidigm AccessArray use LNA groups in their custom sequencing primers to increase stability without using long oligos.

[Ktu]

Good morning everyone,

i'm currently using Miseq for sequencing. But i want to use a custom sequencing primers. If i understand well the company doesn't share the information about GC% neither the Tm for the primers used in the sequencing. So how can we design a sequencing primer. What is the optimal lenght, Tm and GC%?

Thank you for you help

[kcchan]

Ktu, please re-read this thread carefully. The sequencing primers has already been posted by pmiguel. The Illumina provided primers are actually a pool of several different primers, but their characteristics are all fairly similar.

[Ktu]

Originally Posted by kcchan

Ktu, please re-read this thread carefully. The sequencing primers has already been posted by pmiguel. The Illumina provided primers are actually a pool of several different primers, but their characteristics are all fairly similar.

thank you for your answer but i'm using nextera DNA kit for my library and the sequences of the adaptors don't match with the sequencing primers posted by pmiguel.

In the illumina's letter, they give those informations but nothing about primers. Maybe i miss something

Nextera® transposase sequences (FC-121-1031, FC-121-1030)
5’ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG
(a) Read 1 -->
5’ GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG
(d) Read 2 -->

Nextera® Index Kit - PCR primers (FC-121-1012, FC-121-1011)
5’ AATGATACGGCGACCACCGAGATCTACAC[i5]TCGTCGGCAGCGTC
(c) i5 Index read -->
5’ CAAGCAGAAGACGGCATACGAGAT[i7]GTCTCGTGGGCTCGG
<-- i7 Index read (b)

[pmiguel]

Originally Posted by kcchan

Ktu, please re-read this thread carefully. The sequencing primers has already been posted by pmiguel. The Illumina provided primers are actually a pool of several different primers, but their characteristics are all fairly similar.

I may have posted adapter sequences, but for the most part Illumina refuses to reveal their primer sequences. My guess is that is because, like Fluidigm they rely on LNA, or something similar to increase the Tm of their primers for the MiSeq. And someone had sworn them to secrecy about that. (Like that was a condition of their acquiring a license to use that technology.)

That said, the have a technical doc that specifies the maximum temps an annealed primer would be subjected to during cycling. And I don't think there is a down-side to just cranking your Tm as high as possible. Either using LNAs or long primers. Remember this is not like Sanger cycle sequencing -- the primers are not re-annealing during the sequencing process.

--
Phillip

[NWBiotech]

I am using custom primers on a low complexity library. I used positions 18, 19 and 20 to put in the primers. Will the system read the PhiX clones (i.e. are the standard primers mixed into the run mix) or do I need to mix my custom primers into a different position on the block?

[NWBiotech]

Just to clarify, the run with my custome primers did not yield any data (0.4% of clusters passed filter).