Dual indexing construct for MiSeq - where to put the barcode for index2?

Vinz · 02-21-2012, 04:21 AM

We would like to use the second indexing read on the MiSeq for amplicon sequencing (not using the Illumina Custom Amplicon Sequencing). Has anyone successfully designed the adaptors containing index2? Where would you put the barcodes?
I assume "TrueSeq Universal Adapter" from the "Illumina-Customer-Sequence-Letter" would be the sequence to start.
Thanks
Vinz

koadman · 02-22-2012, 04:31 PM

Hi Vinz, I looked into this and the information that Illumina distributes to customers does not contain the full length adapter sequences as far as I know. You might also be interested in this thread where csquared helpfully suggested that the 2nd barcode is inside the other adapter and is primed directly from the flowcell annealing sequence. Reading Illumina's instrument operation manuals appears to confirm this, they suggest there are some chemistry-only dark cycles prior to reading the other index in their standard protocol and these presumably synthesize the bases between the flow cell annealing region and the index region.

However, if you are doing amplicon sequencing you probably have the flexibility to design your own sequencing primers in which case you might be able to follow a design like what Kircher et al have described. That would have the advantage of not wasting 9 cycles of reagents.

Vinz · 02-24-2012, 03:55 AM

Thanks for the link to the paper of Kircher et al. That is indeed helpful.
The way the 2nd barcode is sequenced is nicely described in the Dual Indexing Training for all that have access. It just does not show any sequences.

koadman · 02-24-2012, 08:42 AM

The video in that training is the best explanation I've seen yet of the dual index adapter structure and sequencing process. Interesting that the 2nd index is sequenced while the template molecule is bridged. Thanks Vinz!

ECO · 05-08-2012, 09:08 AM

Hey guys. I'm in the process of making my own Truseq-esque double indexing adapters for MiSeq, and (as you've seen) Illumina will not be forthcoming with the sequences...so I had to figure it out for myself and will share it with you.

I just sequenced a standard single-indexed TruSeq library with the Nextera/Amplicon protocol and the below describes my observations.
It may not help with figuring out the nextera sequences but probably contains some information that will help.

I'm just about to run libraries with my own homebrew double indexing design, will post when/if it turns out well.

ECO · 05-08-2012, 09:10 AM

Oh and page 5 of this document has as nice diagram of the process, so far the chemistry has seemed the same on the MiSeq.

Vinz · 05-10-2012, 11:47 PM

Meanwhile we figured out how to do the dual indexing.
What you have colored green are the 7 bases that are cut with an restriction enzyme and later added again by the 7 dark cycles. We hypothesize the restriction enzyme is DpnI. Recognition site GATC, cutting after GA.
This green sequence should be as it is on the left side of the index. After the index the ACAC is repeated as this matches the 5' end of read 1 sequencing primer and then you go on with the red part...
This works very well for us on the MiSeq for dual indexing with the primers of the cartridge.

gnomers · 06-14-2012, 12:48 PM

Quote:

Originally Posted by ECO

I'm just about to run libraries with my own homebrew double indexing design, will post when/if it turns out well.

By "homebrew," do you mean that you were using custom Read 1, i7 index and Read 2 primers? I am contemplating this strategy for dual indexed amplicons in order to avoid sequencing the constant regions that anneal to the PCR template.

Either way, how did it turn out?

seq_yak · 06-19-2012, 08:22 AM

Quote:

Originally Posted by Vinz

Meanwhile we figured out how to do the dual indexing.
What you have colored green are the 7 bases that are cut with an restriction enzyme and later added again by the 7 dark cycles. We hypothesize the restriction enzyme is DpnI. Recognition site GATC, cutting after GA.
This green sequence should be as it is on the left side of the index. After the index the ACAC is repeated as this matches the 5' end of read 1 sequencing primer and then you go on with the red part...
This works very well for us on the MiSeq for dual indexing with the primers of the cartridge.

I have a library design based on Kirchner et al. as well and I'm planning to run it on the MiSeq. So, did you have to tweak the sample sheet of the MiSeq or did you use the standard Nextera program? Maybe you could even share your sample sheet here?

Thanks,
seq_yak

Vinz · 06-21-2012, 05:56 AM

I have attached a recent sample sheet example.
The sequence of index1 (I7) needs to be reverse complemented of what you ordered as primer. The sequence of index2 is just as the primer was synthesized.
However, this will use the index and read primers of the cartridge. If you want to add your own primers you need to add this information.

TonyBrooks · 06-21-2012, 06:30 AM

Does that mean that using the following primers you could create your own dual indexed amplicon library?

AATGATACGGCGACCACCGAGA{TCTACAC}[i5 index][custom fwd]
CAAGCAGAAGACGGCATACGAGAT[i7 index][custom rev]

Dark cycles in curly brackets{ }
Read 1 with custom fwd
Index i7 with reverse compliment of custom rev
Index i5 same as Illumina's
Read 2 with custom rev

We'd be very interested in this for potential 16s studies as could then feasibly multiplex 96, or even 384 samples in one MiSeq run.

Vinz · 06-21-2012, 06:41 AM

I think this should do it.

seq_yak · 06-21-2012, 07:58 AM

Quote:

Originally Posted by Vinz

I have attached a recent sample sheet example.
The sequence of index1 (I7) needs to be reverse complemented of what you ordered as primer. The sequence of index2 is just as the primer was synthesized.
However, this will use the index and read primers of the cartridge. If you want to add your own primers you need to add this information.

Hi Vinz,

thanks for the input. So, if I understood you right you're able to use the primers of the cartridge and the sample sheet you posted to sequence such a TruSeq-esque library (attached)?

Vinz · 06-21-2012, 08:11 AM

Hi seq_yak,
yes, with the design you attached things should work well. You do not need to add additional primers and with the sample sheet MiSeq should give you nice sequences. Good luck!

pmiguel · 06-25-2012, 04:25 AM

Quote:

Originally Posted by Vinz

Meanwhile we figured out how to do the dual indexing.
What you have colored green are the 7 bases that are cut with an restriction enzyme and later added again by the 7 dark cycles. We hypothesize the restriction enzyme is DpnI. Recognition site GATC, cutting after GA.
This green sequence should be as it is on the left side of the index. After the index the ACAC is repeated as this matches the 5' end of read 1 sequencing primer and then you go on with the red part...

So that would be a hemi-methylated DpnI site? With the flowcell oligo contributing the methylated adenine required for DpnI to cut?

--
Phillip

ECO · 06-27-2012, 10:54 AM

Ok sorry for the delay all...please see attached document for my results/experience. There is a review of standard truseq, the design that I came up with with help from this thread, and an example samplesheet.

Everything comes with my standard disclaimer of "if you don't know what you're doing you WILL mess something up" and...this is based on one successful run of testing...so this is very experimental.

Vinz · 06-28-2012, 08:26 AM

Quote:

Originally Posted by pmiguel

So that would be a hemi-methylated DpnI site? With the flowcell oligo contributing the methylated adenine required for DpnI to cut?

--
Phillip

Hi Philip,
that was the idea. But ECO pointed out that it is probably uracil at some position and some uracil dependant cleavage.

ECO · 06-29-2012, 08:30 AM

I could definitely be wrong though!! It seems crazy to rely on restriction when more specificity would be achieved by uracil dependent cleavage.

The flow cell primer having a uracil at that position is a bit of a clue too...assuming that info is correct.

ThePhotosyntheticMan · 07-03-2012, 09:46 AM

Hi all, I'm new to NGS so approach this post with a healthy dose of skepticism.

I'm in the process of designing PCR primers for a targeted approach to sequencing dual indexed amplicons on the MiSeq. After reading ECO's posts, this is what I've deduced about primer design:

FORWARD PCR PRIMER:
5’ – AATGATACGGCGACCACCGAGA{TCTACAC}nnnnnnnnACACTCTTTCCCTACACGACGCTCTTCCGATCT[Forward primer sequence]
REVERSE PCR PRIMER:
5’ – CAAGCAGAAGACGGCATACGAGATnnnnnnnnGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT[Reverse primer sequence]

...where bold = i5 and i7 indices, respectively.

This confuses me because the sequences following the indices are different from those introduced by the transposon. According to Illumina's Customer Sequence Letter, a targeted PCR strategy that mimics samples prepped using Nextera would have primers that appear as follows:

FORWARD PCR PRIMER:
5' - AATGATACGGCGACCACCGAGA{TCTACAC}nnnnnnnnTCGTCGGCAGCGTCAGATGTGTATAAGAGACAG[Forward primer sequence]
REVERSE PCR PRIMER:
5’ – CAAGCAGAAGACGGCATACGAGATnnnnnnnnGTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG[Reverse primer sequence]

...which would mean Read 1, Index 1, and Read 2 sequencing primers are different, hence my confusion. Will either of these strategies work on the MiSeq? Has the above strategy worked for some folk? My thanks in advance.

TonyBrooks · 07-04-2012, 02:20 AM

I think both should work.
The Illumina read primer is actually a primer mix of several primers. I guess there's probably primers for both sequences in there (TruSeq and Nextera).

However, if you're generating library directly from PCR, why not remove those sequences all together and just use custom read primers. Your custom read primers will just be versions of your fwd and rev primer sequences. You'll also waste less reagent by not sequencing your primers. See my post further up.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How to put header back to .bam file?	eilosei	Bioinformatics	6	02-06-2014 06:22 PM
how to construct phylogenetic tree using SNPs	tianyub836	Bioinformatics	10	05-21-2013 07:27 AM
Tmap: how to use it to construct genetic map with large-scale SNP markers	orctyr	Bioinformatics	0	03-16-2011 02:11 AM
construct whole genome reference for HG19	foxyg	Bioinformatics	3	09-08-2010 02:12 PM
Convert Maq's out put to fasta/fastq?	bea	Bioinformatics	5	06-04-2009 02:18 AM

02-21-2012, 04:21 AM	#1
Vinz Member Location: Germany Join Date: Dec 2010 Posts: 80	Dual indexing construct for MiSeq - where to put the barcode for index2? We would like to use the second indexing read on the MiSeq for amplicon sequencing (not using the Illumina Custom Amplicon Sequencing). Has anyone successfully designed the adaptors containing index2? Where would you put the barcodes? I assume "TrueSeq Universal Adapter" from the "Illumina-Customer-Sequence-Letter" would be the sequence to start. Thanks Vinz

02-22-2012, 04:31 PM	#2
koadman Member Location: Sydney, Australia Join Date: May 2010 Posts: 65	Hi Vinz, I looked into this and the information that Illumina distributes to customers does not contain the full length adapter sequences as far as I know. You might also be interested in this thread where csquared helpfully suggested that the 2nd barcode is inside the other adapter and is primed directly from the flowcell annealing sequence. Reading Illumina's instrument operation manuals appears to confirm this, they suggest there are some chemistry-only dark cycles prior to reading the other index in their standard protocol and these presumably synthesize the bases between the flow cell annealing region and the index region. However, if you are doing amplicon sequencing you probably have the flexibility to design your own sequencing primers in which case you might be able to follow a design like what Kircher et al have described. That would have the advantage of not wasting 9 cycles of reagents.

02-24-2012, 03:55 AM	#3
Vinz Member Location: Germany Join Date: Dec 2010 Posts: 80	Thanks for the link to the paper of Kircher et al. That is indeed helpful. The way the 2nd barcode is sequenced is nicely described in the Dual Indexing Training for all that have access. It just does not show any sequences.

02-24-2012, 08:42 AM	#4
koadman Member Location: Sydney, Australia Join Date: May 2010 Posts: 65	The video in that training is the best explanation I've seen yet of the dual index adapter structure and sequencing process. Interesting that the 2nd index is sequenced while the template molecule is bridged. Thanks Vinz!

05-10-2012, 11:47 PM	#7
Vinz Member Location: Germany Join Date: Dec 2010 Posts: 80	Meanwhile we figured out how to do the dual indexing. What you have colored green are the 7 bases that are cut with an restriction enzyme and later added again by the 7 dark cycles. We hypothesize the restriction enzyme is DpnI. Recognition site GATC, cutting after GA. This green sequence should be as it is on the left side of the index. After the index the ACAC is repeated as this matches the 5' end of read 1 sequencing primer and then you go on with the red part... This works very well for us on the MiSeq for dual indexing with the primers of the cartridge.

06-21-2012, 06:30 AM	#11
TonyBrooks Senior Member Location: London Join Date: Jun 2009 Posts: 298	Does that mean that using the following primers you could create your own dual indexed amplicon library? AATGATACGGCGACCACCGAGA{TCTACAC}[i5 index][custom fwd] CAAGCAGAAGACGGCATACGAGAT[i7 index][custom rev] Dark cycles in curly brackets{ } Read 1 with custom fwd Index i7 with reverse compliment of custom rev Index i5 same as Illumina's Read 2 with custom rev We'd be very interested in this for potential 16s studies as could then feasibly multiplex 96, or even 384 samples in one MiSeq run.

06-21-2012, 06:41 AM	#12
Vinz Member Location: Germany Join Date: Dec 2010 Posts: 80	I think this should do it.

06-21-2012, 08:11 AM	#14
Vinz Member Location: Germany Join Date: Dec 2010 Posts: 80	Hi seq_yak, yes, with the design you attached things should work well. You do not need to add additional primers and with the sample sheet MiSeq should give you nice sequences. Good luck!

06-29-2012, 08:30 AM	#18
ECO --Site Admin-- Location: SF Bay Area, CA, USA Join Date: Oct 2007 Posts: 1,358	I could definitely be wrong though!! It seems crazy to rely on restriction when more specificity would be achieved by uracil dependent cleavage. The flow cell primer having a uracil at that position is a bit of a clue too...assuming that info is correct.

07-03-2012, 09:46 AM	#19
ThePhotosyntheticMan Junior Member Location: North Carolina Join Date: Jul 2012 Posts: 2	Hi all, I'm new to NGS so approach this post with a healthy dose of skepticism. I'm in the process of designing PCR primers for a targeted approach to sequencing dual indexed amplicons on the MiSeq. After reading ECO's posts, this is what I've deduced about primer design: FORWARD PCR PRIMER: 5’ – AATGATACGGCGACCACCGAGA{TCTACAC}nnnnnnnnACACTCTTTCCCTACACGACGCTCTTCCGATCT[Forward primer sequence] REVERSE PCR PRIMER: 5’ – CAAGCAGAAGACGGCATACGAGATnnnnnnnnGTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT[Reverse primer sequence] ...where bold = i5 and i7 indices, respectively. This confuses me because the sequences following the indices are different from those introduced by the transposon. According to Illumina's Customer Sequence Letter, a targeted PCR strategy that mimics samples prepped using Nextera would have primers that appear as follows: FORWARD PCR PRIMER: 5' - AATGATACGGCGACCACCGAGA{TCTACAC}nnnnnnnnTCGTCGGCAGCGTCAGATGTGTATAAGAGACAG[Forward primer sequence] REVERSE PCR PRIMER: 5’ – CAAGCAGAAGACGGCATACGAGATnnnnnnnnGTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG[Reverse primer sequence] ...which would mean Read 1, Index 1, and Read 2 sequencing primers are different, hence my confusion. Will either of these strategies work on the MiSeq? Has the above strategy worked for some folk? My thanks in advance.

07-04-2012, 02:20 AM	#20
TonyBrooks Senior Member Location: London Join Date: Jun 2009 Posts: 298	I think both should work. The Illumina read primer is actually a primer mix of several primers. I guess there's probably primers for both sequences in there (TruSeq and Nextera). However, if you're generating library directly from PCR, why not remove those sequences all together and just use custom read primers. Your custom read primers will just be versions of your fwd and rev primer sequences. You'll also waste less reagent by not sequencing your primers. See my post further up.