[giulianog]

Ion proton

I saw it yesterday at PAG. The instrument will be shipping Q3, 2012. The present chip does two human exomes and has a much larger surface than the Ion Torrent one, they're working on one that will do a complete genome. Present read length 200 bases, it will upgrade after the Ion Torrent.

[aleferna]

So in Q3 2012 they will have a $400k equivalent of a $100k MiSeq?? And therefore in Q1 2013 they will have a $400k equivalent to a $200k GAIIx, and in Q3 2013 a same price equivalent of a HiSeq 2000. Why are they still selling the IonTorrent so expensive? If somebody from Life is reading this, you need to reevaluate your product line and costs. You cannot beat the MiSeq for 2 or 3 chip cycles. You need the HP model, give away the printer and charge for the cartridges, otherwise you won't sell a system for 2 years. By the time you get to beat Illumina they will have such a grip in the market you will never get in.

[SeqAA]

So in Q3 2012 they will have a $400k equivalent of a $100k MiSeq?? And therefore in Q1 2013 they will have a $400k equivalent to a $200k GAIIx, and in Q3 2013 a same price equivalent of a HiSeq 2000. Why are they still selling the IonTorrent so expensive? If somebody from Life is reading this, you need to reevaluate your product line and costs. You cannot beat the MiSeq for 2 or 3 chip cycles. You need the HP model, give away the printer and charge for the cartridges, otherwise you won't sell a system for 2 years. By the time you get to beat Illumina they will have such a grip in the market you will never get in.

You are lucky if you even have a Miseq in hand. It doesn't help they announced enhancements before many even shipped. We need to live in more tangible world where we judge current products, and not just what companies are promising. It's valuable to look at projected capabilities, but let's look at who is producing what right now.

But to be real, your equivalents make no sense.

[krobison]

So in Q3 2012 they will have a $400k equivalent of a $100k MiSeq?? And therefore in Q1 2013 they will have a $400k equivalent to a $200k GAIIx, and in Q3 2013 a same price equivalent of a HiSeq 2000. Why are they still selling the IonTorrent so expensive? If somebody from Life is reading this, you need to reevaluate your product line and costs. You cannot beat the MiSeq for 2 or 3 chip cycles. You need the HP model, give away the printer and charge for the cartridges, otherwise you won't sell a system for 2 years. By the time you get to beat Illumina they will have such a grip in the market you will never get in.

Your comparison is utterly incomprehensible -- you appear to be comparing the list price of MiSeq to the 5-year cost of Ion; do MiSeqs really have no service costs?

IF they meet their spec, the initial Proton would appear to beat MiSeq on performance. Again, IF they launch the second generation chip on time and at spec, then it is a nearly HiSeq class machine for a sub-GAIIx purchase price and a run cost in the neighborhood of a MiSeq.

[Joann]

Ion Proton--some field work in CT

http://www.yaledailynews.com/news/2012/jan/18/yale-gets-new-dna-sequencer/

[TonyBrooks]

So in Q3 2012 they will have a $400k equivalent of a $100k MiSeq?? And therefore in Q1 2013 they will have a $400k equivalent to a $200k GAIIx, and in Q3 2013 a same price equivalent of a HiSeq 2000. Why are they still selling the IonTorrent so expensive? If somebody from Life is reading this, you need to reevaluate your product line and costs. You cannot beat the MiSeq for 2 or 3 chip cycles. You need the HP model, give away the printer and charge for the cartridges, otherwise you won't sell a system for 2 years. By the time you get to beat Illumina they will have such a grip in the market you will never get in.

From what I understand by Q3 2012
MiSeq ($100k) 15m PE reads @ 250bp (~7-7.5Gb) - $500 per run
Ion Proton ($150k) ~5-600m SE reads @ 200(?)bp (>90Gb) - $1000 per run

Ion Proton is more to buy and run but offers >10X output.

[scbaker]

From what I understand by Q3 2012
MiSeq ($100k) 15m PE reads @ 250bp (~7-7.5Gb) - $500 per run
Ion Proton ($150k) ~5-600m SE reads @ 200(?)bp (>90Gb) - $1000 per run

Ion Proton is more to buy and run but offers >10X output.

I think by 3Q12 the Ion Proton will be at 10Gb (with 165M feature Proton 1 chip). It won't be until 1Q13 that they'll have the 100Gb Proton 2 chip.

[DICER]

A miSeq is more over 120 K.... get with the program!

With a Proton 2 chip at 660 million wells AND their paired end technology @ 200 reads, it will be spectacular... especially if you had libraries ready to go. Imagine 8-10 runs in a day.

:0)

[aleferna]

My apologies, I didn't see the 5 year service, so its $250k. Nevertheless, I think it is a strategical mistake to leave the $100k market to illumina for 2013. The Ion Proton sounds nice, but you can no longer think of a PGM vs MiSeq. My point is that, we were forced to buy a Miseq because the PGM+Server+OneTouch price tag was the exactly the same as a MiSeq, I was expecting a $50k PGM and we going to have to pay more than twice. I think few people will choose a PGM over a MiSeq at the same price tag during 2013, specially now that we know the PGM is at the end of the line. Ion is going in the wrong direction, they should have a $20k system with 1Gbp, flood the market their product and get the ROI out of KITs. I don't think they can compete with the high end, in any case who cares about 600Gbp? Those "mainframes" only work in sequencing centers (which are doomed anyway), where you grab a lot of samples and you multiplex them, what about the idea of a small machine that you can use for a single sample... You can either sell 10,000 x $500k machines, or 1M x $50k machines, I say go with the $50k and get the money from reagents... Why do you think printers are so cheap?

[scbaker]

# of reads

Some users (including ourselves) are managing to get >75% live wells on the 316 chip or even higher.

I think this will still be a bead-based system (but smaller beads, wells and spacing), but it's an interesting question!

Nick,

Do you mean that you're getting productive reads from 75% of the wells (~4.5M reads)? That's pretty impressive! That means you should be getting close to 1Gb of sequence from your 316 chips (assuming 200b read length). Is that right?

Shawn

[MrGuy]

I think you have it wrong. The 316 chip is the 100 MB chip. The 318 chip is the one that gets 1 GB. It looks like you are off by an order of magnitude.

... Ok, looks like I should do some math before reiterating the market hype. Assuming you really are getting 75% occupancy AND that can be done on the 318 chip, then the following applies. Correct me if I am wrong:

318 chip
11.3 M features --> 75% live wells
200 bp reads is ~17Gb
400 bp reads is ~34Gb

Assuming this applies to the new chips of
165M features, 75% occupancy
200 bp is ~250Gb
400 bp is ~500Gb

660M features, 75% occupancy
200 bp is ~1000Gb
400 bp is ~2000Gb

Does this even make any sense or am I completely high?

[scbaker]

I think you have it wrong. The 316 chip is the 100 MB chip. The 318 chip is the one that gets 1 GB. It looks like you are off by an order of magnitude.

... Ok, looks like I should do some math before reiterating the market hype. Assuming you really are getting 75% occupancy AND that can be done on the 318 chip, then the following applies. Correct me if I am wrong:

318 chip
11.3 M features --> 75% live wells
200 bp reads is ~17Gb
400 bp reads is ~34Gb

Assuming this applies to the new chips of
165M features, 75% occupancy
200 bp is ~250Gb
400 bp is ~500Gb

660M features, 75% occupancy
200 bp is ~1000Gb
400 bp is ~2000Gb

Does this even make any sense or am I completely high?

Yeah, that's basically my question as well. If people are really getting 75% of the features to produce good reads, then it seems like these chips should be pumping out a lot more data than is generally talked about.

[MrGuy]

A little more math to really mess with you: Ion is saying they can do the genome for $1000. However, is this with the calculation of 200 bp or 400 bp at the less efficient live wells rate? Let's assume 75% live wells... That means the human genome could be done 10-20 at a time on the 660M chip for $50-$100 per genome. And the proton does a run 2 times a day. Max 40 genomes per day. 300 working days per year. That's 12,000 genomes per machine. If BGI bought 138 of these, that would be 1,656,000 genomes per year.

Please, someone check my math. I feel ill.

[Chipper]

A little more math to really mess with you: Ion is saying they can do the genome for $1000. However, is this with the calculation of 200 bp or 400 bp at the less efficient live wells rate? Let's assume 75% live wells... That means the human genome could be done 10-20 at a time on the 660M chip for $50-$100 per genome. And the proton does a run 2 times a day. Max 40 genomes per day. 300 working days per year. That's 12,000 genomes per machine. If BGI bought 138 of these, that would be 1,656,000 genomes per year.

Please, someone check my math. I feel ill.

If we assume 50% usable reads and 200bp read lengths for the 660 M well chip then the $1000 genome would be sequenced at 22x coverage with is rather low, but could qualify as a "complete genome". Probably at the time it will be able to do longer reads (or perhaps better yet overlapping 2x200 bp) witch would give a high quality genome. So "only" ~ 100 000 genomes for BGI in that case. But of course, if the 2x300 bp MiSeq read lengths and cycle times comes together with doubling of the cluster density their current fleet of HiSeqs they could get another 300 000 genomes / year.

[westerman]

A little more math to really mess with you: Ion is saying they can do the genome for $1000. However, is this with the calculation of 200 bp or 400 bp at the less efficient live wells rate? Let's assume 75% live wells... That means the human genome could be done 10-20 at a time on the 660M chip for $50-$100 per genome. And the proton does a run 2 times a day. Max 40 genomes per day. 300 working days per year. That's 12,000 genomes per machine. If BGI bought 138 of these, that would be 1,656,000 genomes per year.

Please, someone check my math. I feel ill.

I haven't looked at this very hard but it does seem like your math is wrong. You say that a 660M feature chip at 200 bp per feature at 75% is 1000GB but when I calculate 660M*.75*200 I get closer to 100GB. All your numbers seem to be off by a factor of 10.

100GB is about 30x human genome coverage. Which is what I would consider the minimum for good variant coverage. So I would be willing to say closer to 2 human genomes per machine per day. 600 per machine per year. BGIing it at 138 machines, say 82,000 genomes per year.

Feel better?