What are the appropriate terms and defintions in sequencing?

[Lluc]

It seems that there is no agreement on the usage of some terms in massively parallel sequencing. For example, the SAM specifications define as "segment" what many would call "read", namely "A contiguous (sub)sequence on a template which is sequenced or assembled." And I understand that their definition of "read" would include the two paired ends of a template sequenced by Illumina, e.g. So, according to SAM specifications, we could say "discordant read", instead of "discordant pair"; and it would not be appropriate to refer to a single end with the term "read". Convenient as it seems to standardize the use of these terms, the truth is that I don't see the terms used in the literature in the way suggested by SAM.

Another problematic term is "insert", when applied to the fragment of DNA between the two sequenced ends of a pair, which may have never been "inserted".

Then, the terms "clonal coverage", "span coverage", and also "physical coverage" have been used with apparently the same meaning.

Lastly, I'm bothered by the flashy qualifier of the "next generation" sequencing. Doesn't it make more sense to say something less ephemeral, like "massively parallel" sequencing?

I welcome any recommendation, reference, guidance or comment on the appropriate usage of these terms.

[Torst]

You are correct, it is a bit of a mess!

The 'insert' is legacy terminology from BAC libraries etc. We should probably say 'fragment size' as it makes it clear that the length INCLUDES the original read(s).

I've never heard those three coverage terms. Usually just here 'coverage', 'depth' or 'depth of coverage'.

"Next-Gen" was ok originally, but many people (including myself) try to use terms like "high-throughput sequencing" (HTS).

[GW_OK]

I think most people are standardizing (or at least they should) on calling Illumina/454/Solid/Ion "second generation sequencing". Pacbio/Nanopore stuff should be "third generation sequencing". We can argue about Helicos but it's a moot point right now.

I too think there needs to be some consensus as to what a "read" is as Illumina counts forward and reverse reads separately for paired-end sequencing in their product literature, while I think they both together should be called a "read".

[scbaker]

I think the terms "second gen" and "third gen" are hopelessly pointless. PacBio and ONT are 'third' because they're single molecule? What about new nanopore technologies that require sample prep (possibly with some amplification)? What would you call GnuBIO? As the number of available platforms and technologies increases, it becomes more and more difficult to put them into (meaningless) categories. I just call them all 'next gen', but HTS probably makes more sense.

[GW_OK]

I contend "high throughput" is meaningless as well.
An ABI 3730 is "high throughput" as compared to an ABI 377, no?
I'd say anything that detects a clonally amplified molecule is 'second' gen.
If it sequences the single molecule itself it is 'third' gen.

But go ahead and call it all 'next gen', because it is easier and it's best to not get bogged down in semantics.

[scbaker]

I contend "high throughput" is meaningless as well.
An ABI 3730 is "high throughput" as compared to an ABI 377, no?
I'd say anything that detects a clonally amplified molecule is 'second' gen.
If it sequences the single molecule itself it is 'third' gen.

But go ahead and call it all 'next gen', because it is easier and it's best to not get bogged down in semantics.

Yeah, that's a really good point. Plus, I think some of the new technologies might focus more on quick, inexpensive sequencing (but not so much on the massive outputs).

[Joann]

Do you think it will get to the point soon where "next gen" is implied and then it would be either clonally amplified (CA) sequencing or single molecule (SM) sequencing?

[mchaisso]

In a draft of a recent paper I tried to coin the term Amplified Cyclic (AC) sequencing for flow cycle sequencing methods (Illumina and IonTorrent), and for single molecule (SM) sequencing, to roll the sequencing as SMS. I got some pushback against coining a new term though, and resorted to Next Generation for AC reads.

cheers,
-mark

Do you think it will get to the point soon where "next gen" is implied and then it would be either clonally amplified (CA) sequencing or single molecule (SM) sequencing?

[Joann]

Thanks for the feedback. I will take this and other related forum-posted topics to the FGED meeting in Toronto next week on standards development.