Friday, July 29, 2011

How Many Toes Does Ion Have Left?

I've previously complained about Ion Torrent's bungling and secrecy when it comes to educating their current and potential future users about key technical information.  I've recently come across yet another botch, one that underscores that attempting to control information in the Internet age only serves to distract from the goal of ensuring access to correct information.
Undaunted by our earlier misfire, I had designed new primers for a new amplicon experiment.  For the experiment, I needed barcoded fusion primers and designed such a set.  My PCR wizard thought it wise to check my designs along with his summer protege, and reported that my B oligo was wrong!!  Horrors!

Now, mistakes happen.  Transpositions of letters bedevil you anytime you transcribe information from a published source, and even copy-and-paste should be handled with care, as it is all too easy to miss a base or few.  I've definitely suffered the pain of committing sequence errors to physical molecules, so constant vigilance is critical.  But, I had triple-checked these (at least!).  How did I get it wrong this time?

So I rechecked the Application Note document which Ion representatives had pointed me at, and discovered to my surprise that my sequence was clearly completely correct.  How had my wizard been deceived?  So I asked him for the document he had consulted, and he sent it to me.  This document clearly had a different 'B' primer sequence.  Other than the updated figure, I could find only one other difference in the printed document -- a cryptic code in the fine print at the end.  Looking at the URLs, the filename for each had an embedded date, which supported the primacy of my document.  But, somehow Mr. Google or Ms. Bing or one of their ilk was able to find the dangerously obsolete file.  Indeed, there appear to be multiple sites hosting it, and while many are just generic document (and probably computer virus) sharing sites, at least one is being presented by what appears to be an authorized Ion Torrent distributor. ¡Ay carumba!

My father has now spent precisely half his life trying to impart his wisdom upon me, and now I have spent a bit over a quarter of my life in a similar pursuit.  We have certainly both found it to be a frustrating and incomplete process, with much of the delivered product seemingly uninspected and unabsorbed.  Still, one lesson he taught me that I at least remember is that it is critical to date one's work!  Even more so when that work is potentially changeable and for the consumption of others!   Trying to sanitize the web of your old documents is a fool's task; ensuring that the currency of a document is readily discernible is a sage's.

I do truly hate the amplicon sequencing application note, as it is 99.99% marketing with a tiny bit of useful information hidden (and much unincluded).  Given that this is potentially a huge market for Ion, they seriously need to come up with a technically-minded amplicon design quick reference to replace it.  Such a document, which would be essentially free of marketing and purely technical in nature, absolutely must have revision dates and notes about which versions of template preparation reagents the document with which the document is compatible. 

One final curiosity.  Comparing the two sequences, there is one striking difference (I would post them, but this is the curiously sensitive issue for Ion on which they take umbrage).  The old "B" sequence ended in TCAG, the all-important key sequence used in their base calling model.  This would suggest they were considering either a paired-end strategy or at least one in which alternative emulsion PCR reagents would allow sequencing in from the other end of amplicons, which has proven useful in the 454 world.  What is odd is that the new "B" primer does not end in a key sequence; the last four nucleotides lack a C (TGAT). So, at least for the moment, Ion is not planning on going down those roads.


Anonymous said...

Adapter B doesn't even exit anymore. That's over 2 revisions old. Your PGM contact is quite dated.

Keith Robison said...

You are correct, I should have called it "P1". Symmetric ids like "A" and "B" are better at sticking in my head.

Anonymous said...

My father told me, never imput to malevolence what can be explained by negligence.
If you just look at the user manuals, they are full of typos and contradictory statements, never mind lacking clear and easy to follow instructions
From this, I conclude that the Ion staff is totally overwhelmed with keeping up with all the users flooding them with questions, an enviable problem for any company.
to put it another way, you would think the author of OMICS!! would be smart enough to know that early adopters are...early adopters; it doesn't matter if it is a new car, a new cell phone or whatever, early adopters are guinea pigs (or, perhaps more appropriately, guinea coli).
In other words, stop whining.
Ion is trying hard; if you don't like the rough edges, wait a year.

Anonymous said...

Re. the previous comment; I hardly think Ion Torrent is in early access anymore. If they are prepared to sell an instrument they expect to be used widely, they should have proper channels to track and update revisions and provide easier access to this information for users.

weasel said...

The lack of clear information is a real problem with the Ion Torrent. The ion community website is disorganised and the idea of having to register and keep non-owners in the dark is really annoying. So far I am very unimpressed with support for the system. Given the choice I'd go for a MiSeq.