Thursday, 28 March 2013

Life Tech beats Illumina to the starting gates

Updated after reading this post on GenomeWeb!

Oxford University and the NHS announced the UK's first next-generation sequencing based test for Cancer. The test uses Life Technologies Ion Torrent PGM and AmpliSeq to assay 46 gene "hotspots", but should increase to around 150 genes in the future. The standard AmpliSeq and analysis software have been significantly tweaked to fit into a clinical diagnostics environment.

Wednesday, 27 March 2013

Finding your next post-doc lab is like getting a date for Saturday night (sort of)

A colleague and I were discussing the woes of post-docs trying to find their next job, or a tenure-track position. My colleague was relating the story of a mutual friend who recently published an excellent first-tier publication and was still having problems finding a junior group leader position. She’d said to me that “If so-and-so could not find a job what was the chance for the rest of us mere mortals” implying that you can only get on with multiple Nature, Science or Cell papers!

Tuesday, 26 March 2013

Imputation of LOH from 1x genome sequence

I would like to move from microarray based genotyping to NGS for copy-number and loss-of-heterozygosity analysis. The copy-number bit should be relatively simple and I hope we’ll have done this by the end of the year, however the LOH analysis is more complex as it requires high quality genotype calls and to get these from sequencing you need 10x coverage. Or do you?

Monday, 25 March 2013

Making NGS greener

Does your lab look like this? We get most of our deliveries on dry-ice shipped from European distribution centres. All the polystyrene and dry-ice are the tip of our energy consumption iceberg. Genomics sciences have as much environmental impact as just about everything else, and perhaps quite a bit more than other disciplines.

Tuesday, 19 March 2013

Money laundering 101 for core facility managers

It's that time of year again. The end of the financial year (for many of us) when most core facility managers get frantic calls asking if work can be billed now but done later. Lets face it this is simply a money laundering exercise but two things irk me when it comes round to this time of year.

Firstly why can't grant agencies simply be more flexible with how money is spent on a grant. If the time lines change and money is left over in one year there should be a reasonable degree of flexibility in moving it to another financial year.

Secondly, why can't PI's keep better track of their money. In some institutions budgets can run into many $100,000s or even $millions. Finding out in the middle of February that you have nearly $100,000 left that needs to be "spent" by March 31st does not give you much time to spend it wisely.

My two most frequently requested money laundering methods are:
  1. Buying reagent kits for use later in the year: a simple one to do as most companies will happily expedite reagents to get a large order. But please don't expect your core manager to spend a week negotiating over price, and do remember reagents have a shelf-life!
  2. Buying services ahead of time: again a simple one to do and remember that the service you asked for can also go out of date. What happens if the service provider stops running your favourite array? What happens if a technology changes and you're committed to what now looks like an exorbitant price?
What damage does the system do to science? From my perspective the biggest problem with the sometimes "use it or lose it" accounting is that we can make rushed, and at worst bad decisions. There is nothing like a tight deadline to force people into taking shortcuts. Perhaps only getting one quote instead of three, or not considering the cost of future commitments.

I was never trained as an accountant but for the past fifteen years I have been responsible for pretty large budgets. Most core managers and group leaders that I know are spending considerable sums of often public money and this last minute pressure is one I am sure we could all do without. A bit of extra financial training wouldn't go amiss for most of us either.

If you are laundering money for someone, or requesting monies to be spent at short notice do think about the impact. I have known one group that were faced with buying a new set of reagents because their early purchase went out of date. The samples were irreplaceable and my recommendation was to buy new kits. It was a painful decision and one we would not have needed to take had we been able to reserve the money for closer to when the samples were actually ready!

Wednesday, 6 March 2013

Derek and Clive live (sort of): Nick Loman chats to Clive Brown at ONT

Nick Loman had a nice cosy chat with Clive Brown at AGBT and tells us all about it on his blog Pathgogens: Genes and Genomes. It reads to me like a very reasonable assessment of why we have heard so little from Oxford Nanopore in the last 12 months. ONT have a very disruptive technology that could be months, years or decades from release, but when it comes it could knock the socks off Illumina and others.

I completely agree with Nicks comments that when ONT deliver the community will be quick to forget the lack of information. They hype is created as much by our own discussions as ONT's announcement followed by silence.

I was surprised to hear Clive say the ONT were caught off-guard with the interest in their AGBT 2012 presentation. No-one in the world could stand up at the biggest sequencing technology meeting their is, tell everyone about MinION and then expect us to nod serenely and go back to MiSeq vs PGM (the battle is clearly being won). We were excited, enthralled and desperate to get our hands on the technology first. No wonder we pesterd and griped, most of us probably felt like we might be missing out not to be in the early-access program (unless you're all in it and I'm the only one who isn't).

I, like Nick, sure hope I'm not on Clive's naughty list!

Although Nick may find himself on it not for his negative comments on ONT's strategy but more for his "I spotted a distinctive bald head" comment!

PS: Derek and Clive Live had a lot more swearing!

The joy of grant funding for NGS

At regular intervals throughput the year I am asked for some help with costing next-gen sequencing experiments in grant proposals. The request usually comes towards the middle of the week leaving only a day or two before the deadline but that's just how we work in academia!

One thing that has struck me over the last five or six years is how much more sequencing a grantee can do when the samples are ready compared to when the grant was written as there is usually a delay of months (even years). If you take a look at the graphs below showing how costs have dropped between Jan '10 to Sept '12 you hopefully get the idea.
Wetterstrand KA. DNA Sequencing Costs: Data from the NHGRI Genome Sequencing Program (GSP) Available at: Accessed [March 6th 2013].

Making research $$$ go further: A grant awarded in 2010 for $500,000 might be spent over 3-5 years. If the bulk of the grant is for sequencing genomes, and the grantee spends some time collecting samples the affect on the number of genomes the grantee can sequence can be huge. The graph below shows that 11 genomes could be sequenced for $500,000 in January 2010, wait six months and this rose to 16 genomes. Wait a year and you could have sequenced 24, two years 65 and if you could hold out for three years you could sequence almost 10x as many genomes for the same $500,000.

Will the NIH, Wellcome and others wake up to this: I've not looked at this before and it is rare to go through such a precipitous drop in the cost of a method. However most pundits are predicting the price to continue to drop and Illumina already announced big increases to capacity. Maybe we should be awarding the sequencing portion of a grant at 50% of what was asked for?

Tuesday, 5 March 2013

What other medical tests could you buy with a $1000 genome

There is lots of discussion about the $1000 genome, what it means, what its impact might be, etc.

First off though deciding what a $1000 genome means seems to be almost impossible given everybody uses different ways to calculate costs. In a soon to be published review article we pointed to the 2011 Sboner et al paper that stated a cost of $6500, and we suggest that today a genome costs about $4000 for 30x coverage. Whether you include instrument amortisation as well as data analysis costs makes a huge difference to the final figure. Is it really closer to $25,000?

Even so a "$1000 genome" can be compared to current medical tests and in this post I've put together a list of ten common tests ranging in cost as comparators. I was surprised at the costs of some tests and can see even NGS based tests coming in at well under $100 (about the same as an ultrasound), and perhaps even $10 in the future (at the same as a pregnancy test).

The price of medical tests:
 Robotic surgery to remove prostate £10,000
Standard course of chemotherapy £5000-10,000
BRCA test or Oncotype DX £2,800
Cancer gene sequencing (all exons) £1000
Overnight stay in Hospital £400
CT scan £400
MRI scan £300
Endoscopic biopsy£350
Muscle Biopsy £200
Ultrasound £100
Skin biopsy £50
X-ray (complex) £25
Full biochemistry profile £20
Full microbiology profile £15
Pregnancy test £7.50
Glucose test £5

Will consumer genomics make NGS even cheaper: If people respond to genomics and ask for it from their healthcare providers then costs could fall. A multiplex PCR that amplifies several cancer genes could be very cheap to run if the numbers are high and restricted to genes with relatively simple interpretations.

Can we do cancer NGS testing for as little as a pregnancy test?

PS: the costs of these tests come from a variety of sources but all are likely to suffer the same problem of how those costs were derived. Hopefully this is more Bramleys to Cox's rather than apples to oranges.

Monday, 4 March 2013

Oxford Nanopore chip announced!


I had a very enjoyable trip to the Science Museum in London this weekend and whilst there was amazed to see an Oxford Nanopore DNA sequencing chip on display.

The word on the street after ONT's AGBT 2012 extravaganza has been more like a quiet grumble about progress, the hype has certainly not been lived up to, now that quiet grumble is beginning to get noisier! Everyone was hoping for their MinION USB sequencer and the Star-Trekesque possibility of truly personal genome sequencing. Although even Jim McCoy didn't have that on his tricorder.

There is not even the tiniest hint of what is being done with academic collaborators, no posters, no papers, no seminars. Now all we can do is ponder on why the technology has not been launched as planned. Surely it was not pure hype to whip up interest in their last round of funding? Perhaps the technology has suffered issues, there is a lot going into what ONT and others are trying to do; hardware, software, enzymes and just about everything else all need to be designed and tested and any one of these could stop the system from working. Legal wrangling is almost certainly an issue with a huge number of patents in this field and companies like Illumina working on their own Nanopore technology.

So here it is: The blurb reads "Oxford Nanopore chip. Tiny pores 10,000 times smaller than a human hair sit in microscopic holes covering the surface of this speedy chip. DNA is read electronically as it zips through each pore, generating DNA sequence data."

Photo of ONT Nanopore chip in the "Who Am I" gallery at the Science Museum, London.

Sorry the picture and text are so crummy.

I'm sure we'll see the real thing soon, hopefully long before the 23rd century!

PS: The ONT chip is next to an ABI SOLiD instrument. Don't read to much into it being next to a technology that is already obsolete !