amplicon seq, 2 step, nextera indices vs homebrew

[thermophile]

Hi all

Currently we are sequencing 16s amplicons on our MiSeq using the Caprasso primers. I have people asking to do a variety of non-16s amplicons and will likely use a 2step PCR approach to minimize the primer costs. So now I'm trying to decide if we should use the nextera index kits (costly) or make our own primers (probably based on the dual index 16S primers from the Schloss lab).

Other than the cost, what should I consider to compare these two approaches? I plan on using Phusion polymerase in both steps if using the homebrew index approach.

thanks

[bunce]

Hi Thermophile, if you have to go down the 2-step PCR approach then consider the implications (for indexing) in these two papers;

http://nar.oxfordjournals.org/content/early/2015/02/16/nar.gkv107.full

http://onlinelibrary.wiley.com/doi/10.1111/1755-0998.12402/abstract

Saving a few dollars on primers can quickly become a false economy when you start dealing with contamination and index jumping.

To avoid 2 rounds of PCR one approach is to use index+primer then ligate on (PCR-free) the adapters - we are using this protocol and it is working well: http://www.ncbi.nlm.nih.gov/pubmed/21431776

good luck with your work flows, Cheers Mike

[thermophile]

Thanks Mike. Any wetlab tips/tricks you use beyond what the paper specifies?

[bunce]

as for tips - we generate 1-2ug of amplicon (crude calculation by nano drop) , ligate adapters at a ratio of 3:1 (adapter:amplicon).... we then Pippin Prep the library as the biggest hassle is getting rid of adapters. Then we qPCR for the library quant.

Overall the protocol took 2-3 goes to optimise... but seems to be working well after a number of successful runs.

We use a combination of 1-step fusion amplicons (with custom primers) and a ligation (primers + indexes). The choice depends on the application. Both are designed not to need a 2nd round PCR (which can complicate data fidelity). Cheers Mike

[kmcarr]

Hi Thermophile, if you have to go down the 2-step PCR approach then consider the implications (for indexing) in these two papers;

http://nar.oxfordjournals.org/conten.../18/nar.gkv107

http://onlinelibrary.wiley.com/doi/1...12402/abstract

Saving a few dollars on primers can quickly become a false economy when you start dealing with contamination and index jumping.

To avoid 2 rounds of PCR one approach is to use index+primer then ligate on (PCR-free) the adapters - we are using this protocol and it is working well: http://www.ncbi.nlm.nih.gov/pubmed/21431776

good luck with your work flows, Cheers Mike

Mike,

The first two links in your post don't work, the link URLs have the elipses in them. Could you fix them up.

Thanks

[bunce]

Mike,

The first two links in your post don't work, the link URLs have the elipses in them. Could you fix them up.

Thanks

Hi kmcarr - thanks of the heads-up have corrected them (they used to work!) - the elipes are a truncation that SeqAnswers uses - the refs are;
1)Phillipe et al 2015. Nucleic Acids Research. "Accurate multiplexing and filtering for high-throughput amplicon-sequencing
2) Schnell et al. 2015. Molecular Ecology Resources. "Tag jumps illuminated – reducing sequence-to-sample misidentifications in metabarcoding studies"

One of the authors of the first of these papers posted it on SeqAnswers a month or so back - looking for comment (see http://seqanswers.com/forums/showthread.php?t=50927) - but (disappointingly) there was no discussion of this issue. Cheers Mike

[thermophile]

reviving this old thread because I think I'm about to start using this protocol. A_adapter_b is the old PE adapter correct? The flowcells still have that sequence? Or have you switched that adapter for one of the newer ones? I will be ordering a suite of these primers because I'll be adding the indexes as well as adapters through this ligation process.

Second, in the ligation they use 5ul 2000 U/ul Illumina ligase? Is that a typo? 10000 U per reaction?

[bunce]

Hi Thermophile, #1) In the adaptors we use the Nextera sequences - so yes from memory we substituted out the old PE one when designing this workflow (I would need to do some digging to be compleltely sure of this - let me know if you need me to do this).

#2) This is OK - it is nearly 1ug of DNA you are ligating (as a pool). Works out to be ~20$ worth of ligase per library. I don't think we have tried to reduce this (but again I could ask some people in the lab)

[thermophile]

Thanks for the reply. I'm actually not going to pursue this right now. I was wanting to do it as a way to avoid ordering barcoded primers for multiplexing 100+samples per run. The ligation cost is way more than primer cost for this number of samples. But I appreciate all your help!

[BioKiwi]

To avoid 2 rounds of PCR one approach is to use index+primer then ligate on (PCR-free) the adapters - we are using this protocol and it is working well: http://www.ncbi.nlm.nih.gov/pubmed/21431776

good luck with your work flows, Cheers Mike

Hi Mike

I wonder using index+primer approach if index is sequenced at the start of red1 or R2.

[bunce]

Hi BioKiwi - if your indexes are read 'in-line' then (separate) indexing reads are not needed. For most of our amplicon workflows we do not carry out indexing, instead we deconvolute the data based on index-primer sequences after the run is complete. Cheers, Mike

[BioKiwi]

Thanks for reply Mike. I am not clear at the start of which read (R1, R2 or both) the inline barcode will be sequenced, if inline barcode is added to 5’ end of forward primer. The structure of amplicon after adapter ligation would be as following:

Amplicon:
Inline barcode-target specific F primer/target regioooooooon/target specific R primer (upper strand)
Inline barcode-target specific F primer/target regioooooooon/target specific R primer (lower strand/reverse complement)

Amplicon after adapter ligation:
P5/inlinebarcode-target specific F primer/target regioooooooon/target specific R primer/P7 (upper strand)
P7-inlinebarcode-target specific F primer/target regioooooooon/target specific R primer/P5 (lower strand)

Upper strand read1: inlinebarcode-target specific F primer/target regioooooooon
Lower strand read1: target specific R primer/target regioooooooon

It seems to me that inline barcode could end up at the start of R1 or R2 depending on if upper or lower strand is sequenced or am I missing something?

[bunce]

Hi Biokiwi.... it kind of depends on how long your amplicon is.

We do a fair bit of single-end sequencing (amplicons ~200bp). Here the setup is (post ligation.
P5-RD1-index1-Fprimer..........seq........Rprimer-index2-RD2-P7
read 1-|----------------------------------------------------------------->
So this reads both indexes (i.e sequences through the 'back end')

Alternatively if your insert is long (e.g. 450 bp) can do a paired end read (i.e. sequencing from RD1 and RD2) where you stitch the reads and then sort on both index 1 and index 2.

Is this any clearer? - might be able to help a bit more on the phone if you want to chat - (you should be able to find my contact detail online). Cheers Mike

[nucacidhunter]

Thanks for reply Mike. I am not clear at the start of which read (R1, R2 or both) the inline barcode will be sequenced, if inline barcode is added to 5’ end of forward primer. The structure of amplicon after adapter ligation would be as following:

Amplicon:
Inline barcode-target specific F primer/target regioooooooon/target specific R primer (upper strand)
Inline barcode-target specific F primer/target regioooooooon/target specific R primer (lower strand/reverse complement)

Amplicon after adapter ligation:
P5/inlinebarcode-target specific F primer/target regioooooooon/target specific R primer/P7 (upper strand)
P7-inlinebarcode-target specific F primer/target regioooooooon/target specific R primer/P5 (lower strand)

Upper strand read1: inlinebarcode-target specific F primer/target regioooooooon
Lower strand read1: target specific R primer/target regioooooooon

It seems to me that inline barcode could end up at the start of R1 or R2 depending on if upper or lower strand is sequenced or am I missing something?

You are right. The library is non-stranded and both strands are able to hybridize to flow cell and sequenced.