A first look at Illumina’s new NextSeq 500

[Brian Bushnell]

I'm not sure; I think it's probably fine for quantification unless there's some bias issue, which I have not looked into. I wouldn't want to use it for variant-calling, particularly because a lot of the errors seem like systematic errors that cannot be overcome simply by sequencing deeper. We do use it for multiplexed single cells, because the NextSeq platform has shown lower rates of cross-talk than HiSeq or MiSeq and single-cell sequencing is greatly affected by even low levels of cross-talk. Also, I understand NextSeq is cheaper per base. But certainly, I would avoid the NextSeq (and HiSeq 3000/4000 which I suspect are similar) when possible, if you have access to Illumina's high quality platforms (HiSeq 2000/2500 or MiSeq).

[GenoMax]

Perhaps what you are observing is differences in bcl2fastq v.1.8.4 and 2.18.x?

bcl2fastq v.2.x is required for processing data from NextSeq and HiSeq 3000/4000. It can be used to process data from all current Illumina sequencers. It does binned quality for reads as I recall.

Is your data processed with the same version of bcl2fastq in all cases or was 2500 data processed using bcl2fastq v.1.8.4?

[Brian Bushnell]

I'm not really sure. The HiSeq quality scores are not binned, though. I'm going to talk to the person who manages the Illumina software versions after gathering some more evidence, because we probably will want to roll back to an earlier version, once it's clear which earlier version was better.

Also, does have experience with 3rd-party Illumina base-callers?

Edit: We are using 2.16 for NextSeq and 1.8.4 for everything else.

[GenoMax]

I'm not really sure. The HiSeq quality scores are not binned, though.

That probably means they are using the older bcl2fastq (or CASAVA) v.1.8.4.

I'm going to talk to the person who manages the Illumina software versions after gathering some more evidence, because we probably will want to roll back to an earlier version, once it's clear which earlier version was better.

Also, does have experience with 3rd-party Illumina base-callers?

That is NOT an option for NextSeq and HiSeq 3000/4000 which require bcl2fastq v.2.1x for conversion. Perhaps you can ask the person in charge to reprocess HiSeq 2500 data using bcl2fastq v.2.18.

I don't know if there are any 3rd party callers for new data.

[AllSeq]

But certainly, I would avoid the NextSeq (and HiSeq 3000/4000 which I suspect are similar) when possible, if you have access to Illumina's high quality platforms (HiSeq 2000/2500 or MiSeq).

Why would the NextSeq and HiSeq 3000/4000 be similar? The use different chemistries and different flow cells. Wouldn't the 3000/4000 be most similar to the HiSeq X? (Or did you just mean they're similar in that they're both bad platforms, but for different reasons?)

[GenoMax]

Why would the NextSeq and HiSeq 3000/4000 be similar? The use different chemistries and different flow cells. Wouldn't the 3000/4000 be most similar to the HiSeq X? (Or did you just mean they're similar in that they're both bad platforms, but for different reasons?)

I think it comes down to the bcl2fastq version used for data processing (binned q-scores) for NextSeq and HiSeq 4000.

Hopefully @Brian will have some clarification once he has chased that down that information.

[Brian Bushnell]

Why would the NextSeq and HiSeq 3000/4000 be similar? The use different chemistries and different flow cells. Wouldn't the 3000/4000 be most similar to the HiSeq X? (Or did you just mean they're similar in that they're both bad platforms, but for different reasons?)

They use 2-color chemistry (IIRC). I don't know if the problem is the chemistry, the optics, or the software; but if it's the software, I'd expect the 3000/4000 to be more similar to the NextSeq than the 2500. Also, I've only looked at a single sample of HiSeq 4000 data, but the quality was low; similar to the NextSeq. Since I've seen both good and bad data from the same NextSeq machine, it's obviously possible to produce good data with 2-color chemistry and NextSeq optics. It would be nice if this was all a software issue.

[AllSeq]

The NextSeq uses 2 color chemistry, but the 3000/4000 uses the 'standard' 4 color chemistry (with patterned flow cells, just like the X).

[Brian Bushnell]

The NextSeq uses 2 color chemistry, but the 3000/4000 uses the 'standard' 4 color chemistry (with patterned flow cells, just like the X).

Ah, my mistake.

[Michal2213]

NextSeq suitable for allele-specific analysis?

Do you think that NextSeq would be suitable for allele-specific analysis? I am using mouse cells with hybrid genome and sort the reads belonging to different alleles based on SNP content. So far I was using HiSeq2000 which worked well. With NextSeq I would get the data several times faster but having read this whole thread I am not sure whether the NextSeq data quality will be good enough.

[Brian Bushnell]

If the cost is reasonably similar, I'd go with HiSeq 2000. But perhaps you can get a sample of NextSeq data for your project, on a library you already ran on the HiSeq, to compare without committing yourself?

[Michal2213]

NextSeq suitable for allele-specific analysis?

If the cost is reasonably similar, I'd go with HiSeq 2000. But perhaps you can get a sample of NextSeq data for your project, on a library you already ran on the HiSeq, to compare without committing yourself?

Thank you for reply! I would do the same but with NextSeq we are getting data within a week as we have direct access and with HiSeq2000 in the sequencing core one need to wait in many cases more than a month. I guess we will need to go for sequencing the same library on both libraries as you suggested as well.

[dsobral]

Just to add our 2 cents.

We have a MiSeq running since 2013, and after some hickups we're now stable with it and reasonably happy.

We just recently installed a NextSeq500 and our first tests are not great. Q30 is >80%, but there are many low quality bases (constantly Q=14 "/"), and the worst part is that instead of being towards the end, they seem a bit randomly distributed. When comparing PhiX in a 2x150bp NextSeq with a 2x250bp MiSeq, after alignment I see a 0.2-0.3% error rate with MiSeq and 0.9-1% error rate with NextSeq (1M sampled reads). In the "randomly" distributed Q=14 bases I seem to notice more A to T transitions, but I didn't have time to gather more systematic statistics... If I do quality trim on the MiSeq I can easily get higher quality data, with the NextSeq since its randomly distributed is harder...

We've complained to the Illumina people, let's see what they say...

[Brian Bushnell]

We've complained to the Illumina people, let's see what they say...

I don't believe that Illumina officially admits that NextSeq is lower quality than HiSeq 2500 or MiSeq. Error rate is not part of the platform specification, just quality scores.

[alexhaj]

Good work Brian! I am skeptical about your claim that NextSeq has less crossover than HiSeq or Miseq however. Would you please provide some data to back this up. And if that is the case maybe it's simply because demultiplexing is being done by CASAVA v2 on NextSeq and CASAVA v1 on HiSeq and Miseq. What if you did demultiplexing yourself, taking into account Quality scores (which I assume is not typically done).