I found a paper from Pinto and Raskin 2012 where they have modified the archaeal primers but not for Illumina. Anyone have tired it?

Could you find your answer? Because now I have your problem.

@Akram: I decided to use the primers amplyifying roughly 300 bp for Illumina sequencing. Maybe this is useful for you:

Arc519F: CAGCMGCCGCGGTAA (Overeas et al., 1997, Teske and Sorensen 2007, Wang and Qian 2009)

Arc806R: GGACTACVSGGGTATCTAAT (Raskin et al 1994, Takai and Horikosh 2000)

Thank you very much for your reply. Could you please give me the PCR condition for your setup?

Did you publish a paper about Archaea?

@Akram: Not yet but soon :D My PCR program was set to the needs of KAPA HiFi polymerase. SO I would suggest to run the conditions based on your polymerase. Generally you can run these primers at 52-55°C annealing temperatures and annealing time can be around 20-30s. Amplification time an be40 s as the product is only 300bp. You can have a final elongation of amplification for 5 mins after 28-30 cycles. Hope it helps. Cheers!

Dear Ajinkya Kulkarni

Thank you very much for your valuable help.

Good luck.

Hi!

my group used 516F and 915R for the illumina sequencing, and we noticed the reverse read quanlity is always bad.

How is your primers? could you updated how is your result looks like?

Best regards!

Tong

@Tong Liu: Can you elaborate how or what is causing the reverse reads to be bad? I know that in Illumina technology, we always obtain generally reads with lower phred scores on checking quality of reverse reads. For some reason (probably assay exhaustion) the reverse reads are generally having bad quality especially at the 3 ' ends. Hope this helps.

Hi!

We haven't found out why the quality of reverse reads are bad. but we noticed only the archaea primer has this problem, but my bacteria primer.

That is why I am looking for a better primer or alternate methods.

In the attachment is the reverse reads the quality of one of my recent sample looks like. As you can see the q value drops very quickly, and I barely make the paired-end overlapped with each other.

So, I am wondering how is your quality result.

Tong.

Hi,

Oh yeah seems the quality drop is really dramatic.

Did you ask the company as to how good the internal quality check was (I think Illumina assays sequence a small stretch of DNA, like an internal standard, after each run to see if the quality was affected)?

Unfortunately the MiSeq assay we tried failed completely due to some sequencer related issues and our barcoding approach was more of a problem that time so we moved to HiSeq runs (150 bp * 2),

I will attach the read quality of that if you would like to check it out.

You can see that in our case we also have generally poor read quality on the R2 side (last 30 bases or so).

Nonetheless a quick search online also showed me that people find it normal that Q score is between 25-30 (which doesnt reject many sequences on filtering downstream) for R2 reads and they seem to give no reason as to why it happens but say that its normal to observe this (biostars, seqanswers etc).

You can see our R2 scores were generally closing in at around ~27 but with huge deviations.

I would definitely contact the company and ask them the reason behind this.

You seem to generally have bad Q scores about after 150 bp which is not so nice.

I dont know how you added the Illumina adapters to your amplicons, DNA etc.

We do a blunt a ligation which means that our amplicons can be oriented with either forward or reverse primer after the forward sequencing primer region.

Based on this I dont think that the primer would be a problem. If that was the case then we would see bad quality reads (due to lack of heterogeneity in reverse primer, maybe) in the R1 file too.

But cant say much as I dont know how you prepared your library.

Hope this helps.

Regards,

Ajinkya

Hi!

Thanks a lot!

I will bring this and discuss in my group and also ask the sequencing company.

Tong

Hi everyone!

I finally got the reply from the sequencing company our group used. I think there is still some specific point unclear, but it most likely is because of the degradation of enzymes

Here is what they replied:

" Hej Tong!

Just like you say, the read2 of Illumina runs almost always have a worse quality than the first read. We see this especially for the longer MiSeq-runs.

I think that the main problem is that the enzyme starts degrading the longer the run proceeds. For most instruments chemistry upgrades have improved in this matter, but not for the MiSeq.

Also, if the library is short and we start reading in to the adapters, the quality might decrease. For amplicon samples, we always add extra PhiX to increase the complexity of the library sequenced. In some cases I think that the quality might be increased by adjusting the amount of PhiX.

The specification for the Q30-value of a MiSeq PE300 run is that at least 75% of the bases should have a quality score of 30 in mean over both reads. However, in most cases we see (like in your case) that the value is much higher for the first read than for the second read.

Just like you say, there might be several reasons for this problem. However, I think that the main problem is the problems with the enzyme mentioned above.

I hope this helps!"

Hi Tong,

Thanks for the update!

I think what might help also is that I am now trying a HiSeq 2 x 300 bp run. Apparently it is better than the MiSeq run (quoting the company we work with). Nonetheless I can post an update once I have the result of such a run (in a month or so).

Then you could compare it and maybe switch to this chemistry rather than the MiSeq one.

Have a nice day.

Regards,

Ajinkya

Hi Tong,

A brief update after our last conversation.

We just conducted an Illumina HiSeq run 2*300 bp with GATC Biotech. We did two library runs and we had no issue whatsoever with the quality of the reads.

The amplicon size with the barcodes (8 bp on each side, without the adapter sequences) was roughly 500 bp. I hope this helps and maybe you can also give that HiSeq chemistry a try since the company you are working with mentioned that the MiSeq chemistry has not been improved.

Wish you a nice Sunday.

Regards,

Ajinkya

I would like to chime in if possible.

Indeed most of researchers are still using Miseq, for amplicon sequencing. However late improvements in Hiseq kits have managed it to create 250 and 300 bp (rapid run mode SBS v2 kit).

We sequenced fungi amplicon using 2 x 250 Hiseq, no coverage issue and no read quality dropping. Today definitely you should consider Hiseq, especially if you need higher coverage to detect minor abundant taxa.

Best,

Filipe Pereira Matteoli : I can recommend your suggestion too as we have also shifted to 2x250 bp and 2x150bp HiSeq runs for our datasets.

We did mcrA gene sequencing and 16S archaeal sequencing without much trouble respectively. However due to primer issues we only used orientated forward reads in the latter case!

Ajinkya Kulkarni as we kind of started this topic here, today i've received Ion S5 SE600 as an option for amplicon sequencing from a facility. However, i am unsure about this option, have you ever worked with data from this platform, does regular pipelines apply to them?