Would you give your right arm for a 1st author Nature paper?

This is a slightly wacky post but I did warn you I'd go off on tangents sometimes!

My new bike, multi-tool comparisons and "how fast can you unfold a Brompton?"

My old bike was stolen back in March. We had gone to the campus bar to celebrate the publication of two papers my group had been involved with; the first on Breast cancer in Nature and the second on using sequencing of ctDNA to identify and monitor cancer mutations in ScienceTranslational Medicine. I was only in the pub for about 40 minutes when my bike was stolen.

When I was walking back to the lab with one of my colleagues we got into a conversation about whether “I’d rather have my bike than the papers”! I loved my bike but perhaps the two papers were worth it?

What do you think, what would you value the papers at and what would you be willing to give up for them? Would it matter where you were in the authorship list and does joint 1^st author status dilute the value?

Is any paper literally worth an arm and a leg?

My new bike: Six months later and I finally have a shiny new Brompton. Six speed and dyno-hub lighting with a beautiful raw lacquer finish (just look at those golden welds) and a stunning Brooks saddle. Incidentally Brooks saddles are the oldest bike components still being manufactured and are pretty much as designed back in 1896!

Multi-tool comparison: I has asked my wfe toe get me a multi-tool to go with my new Brompton for my birthday but was not sure which one to buy. There are lots of reviews but I could not find anywhere to do a direct side-by-side comparison of different features. So I did one myself and thought I’d share it with anyone that can find it. The file is available on GoogleDocs.

 After much deliberation I went for the Topeak Alien II. Mainly because it had the basic tools plus a pedal wrench, knife and bottle opener. I think I could design the perfect tool now as about the only thing missing on the Alien II are the integrated tyre levers from the Hexus II. The perfect tool is out there somewhere! I thought I'd also give a runners up prize to Crank Bros, Lezyne's Carbon 10 and Topeaks Hexus II.

Topeak Alien II

Crank Bros, Lezyne Carbon and Hexus II (with integrated tyre levers)

My wife though I was crazy doing this, but she was watching trash on the TV anyway.

How fast can you fold a Brompton: There is an annual Brompton world championships. The 7th was held on in August at the Bike Blenheim Palace Festival Of Cycling. 682 riders took part with the winner completing the course in just over 20 minutes 17 seconds. The real fun is in the unfolding race where the winner was ready to ride in just 8.35 seconds while the average of the 70 competitors was only just over 16.5 seconds.

Alan Morris was "Best dressed Male" and Audrey Shannon took the "Best dressed female".

Sounds like fun, keep an eye out here if you are a Brompton owner and want to take part next year.

HuID-seq follow-up

A few weeks ago I wrote about an idea I had been thinking about for a while for using SNPs as additional content in capture or amplicon assays to provide patient identification at the same time as generating test data. I have many responses from others that would like to do this and the question I have been asked most is which SNPs should we be using.

I’m afraid I can’t answer that question and think a community discussion would be the best way forwards although I am not sure how best to kick-start that discussion. If you have any ideas do let me know! I’ll ask CRUK’s StratMed board for starters.

One person also pointed to a couple of papers and Sequenom’s iPlex sample ID panel. The assay runs a 52SNP multiplex PCR on the MassArray sysytem. I think the biggest thing missing from the SPIA assay on Sequenom is the lack of coding SNP information. I suggested in the older post that using coding SNPs would allow the assay to be more widely used in research settings and even in future clinical RNA-seq based gene expression panels. PAm50-seq perhaps?

I'm getting some designs done to test different methods in the lab. Feel free to send me suggestions on what criteria should be considered when choosing SNPs.



Paper 1:
Sanchez et al; A multiplex assay with 52 single nucleotide polymorphisms for human identification. The authors are members fo the EU SNP for ID project they recommended 52 SNP markers in the publication where they describe their choice of 52 SNP markers and development and validation of a multiplex PCR assay. The project webist has a very nice tool to visualise SNP distribution across different populations.

Their SNP selction criteria were;

i) Maximum 120bp amplicon size
ii) Minimum 0.28 MAF in one and 0.17 in at least three populations
iii) Random distribution of SNPs
iv) 100kb distance from neighboring marker SNPs or genes
v) flanking DNA sequence reliably reported and free from interfering poly- morphisms, such as nucleotide substitutions in potential primer binding sites.

They analysed SNPs using single-base extension (SBE) with ABI SNaPshot kits, products were detected on a capillary sequencer and analysed with GeneScan software. 


Paper 2: Demichelis et al; SNP panel identification assay (SPIA): a genetic- based assay for the identification of cell lines. Used cell lines run on Affymetrix SNP genotyping arrays to identify candiate SNPs. They commented in their paper that they would have preferred to use whole genome genotype data from a larger number of samples to rank SNPs that best distinguish sampes and using an iterative design could define the “most accurate and parsimonious panel [] and how many [SNPs] are needed”. They used cell lines typed on Affymetrix XBba 50k arrays.

Their SNP selection criteria were;

i) SNPs have assigned rs identifier
ii) SNPs are not located in intronic regions
iii) SNPs are also represented on the 10 K Affymetrix oligonucleotide array.

Their iterative testing generated an average of 30-40 SNPs as the optimum. They developed a SPIA test panel to run on Sequenom’s MassArray which uses multiplex PCR as the method to target specific genomic loci. The SPIA analysis tool is written in R and available here. Their conclusion was that any 40 SNPs from their top 100 would produce a good SNP panel for Human identification but that more SNPs equals more confidence! They also showed that a SNP-ID panel can be used to monitor genetic drift of cell lines during passaging.

A new blog to watch for NGS wanabees

I wanted to point my readers to BiteSizeBio as I have been writing for them for about a year now and my latest post has just been published on the NGS channel, A Short History of Sequencing Part 1.

BiteSizeBio offers; "brain food for biologists". It offers short articles on all things biological from Analytical Chemistry and Bioinformatics to Statistics and Technologies and Techniques. There are tips on how to survive in the lab and fun stuff as well.

New to BiteSizeBio are their blog channels dedicated to particular topics. My latest post is on the Next Generation Sequencing channel, there are others for Cloning & Expression, Microscopy & Imaging and Writing / Publishing / Presenting.

I'd encourage you to also get involved by contributing your own articles or by submitting a tip to make life easier in the lab.

I started writing my blog and the articles for BiteSizeBio because I wanted to do more writing and most of it is done in my spare time at home or during lunch at work. I hope it is helping when it comes to writing or reviewing papers for work, but even if it does not help there it is very enjoyable.

Marketing fun for NGS

There has been a battle going on between Life Technologies and Illumina for NGS supremacy, especially over the personal sequencer market. I am sure you saw the "PGM vs MiSeq" videos on YouTube (1, 2 & 3)? If you didn't they are worth a look.

However I'm sorry to say Illumina still don't gift wrap!

Illumina and Life will always spend a fair proportion of income on marketing. In their 2011 annual reports they spent $261M and $1023M respectively on "selling, general and administrative" which I guess includes marketing somewhere. As a percentage of revenue this is 24% for Illumina and 29% for Life Technologies.

Where is the money being spent today: two recent marketing activities caught my eye, I don't prefer one over the other and they both target very different audiences. Both companies are doing much more than these two examples.

Illumina MiSeq grant:  Illumina are hoping to tempt potential customers by awarding three grants to use the MiSeq system. Awardees will receive a MiSeq, 10 sequencing kits, a sample prep kit (Nextera or TSCA) and access to BaseSpace. It does not matter where you are based or if you have experience with NGs, just submit your idea and keep your fingers crossed.

Ion Torrent's Mini tour: The Ion Torrent bus is still touring the USA, what better to tour the UK than a Mini (shame their German now). The tour is coming to my Institute next week and due to the quick run time of the PGM you can watch a run being setup, listen to a seminar and talk to the sales team, have lunch (I am not sure if lunch is included) and see the results from the run. All in about 2.5 hours. It should be fun and I hope to see some of you there. Register for the 24th of September visit to the CRI.

• 12.30  Meet the PGM in the back seat of a Mini
• 12.45  Start the PGM run 
• 13.00  "Enjoy" a technical seminar, no heckling about MiSeq or GS Junior please
• 15.00  See the results from the PGM run

I am not endorsing either of these particular marketing events. I hope you like them though because we are all paying for them with the instruments and kits we are buying! Think about how much of the fun at AGBT is sponsored by PacBio (Gold), Agilent & Roche (Silver) and Caliper, Illumina, Complete Genomics, NuGen, RainDance and Advanced Analytical (Bronze). Life Tech were sleeping over in the bus and sneaking into the bar!

PS: If Ion want to get rid of a Mini after the event I need a new car. I'd prefer them to follow my lead on a BMW R1200GS motorbike fitted out as a mobile sequencing lab though!

Illumina aquires BlueGnome

Illumina just announced that they have aquired the Cambridge based BlueGnome, the company produces their own arrays for genetic screening and the BlueFuse analysis package.

BlueGnome formed in 2002 to commercialise the BlueFuse microarray analysis software. This was the first software I know of that used Bayesian algorithms to generate intensity values for each spot on the array. It worked very well and I was interviewd for a job selling it way back in about 2003 (I did not get the job).

In 2006 BlueGnome launched CytoChip for genetic screening and according to company stats over 100,000 have been run in 40 countries. In 2009 the first BlueGnome baby was born (I am not joking)! BlueGnome had developed a 24Sure, preimplantation genetic screening (PGS) test to screen for all 24 chromosomes to enable selection of normal chromosome complement eggs. The technology allowed a woman to have her 14th and final IVF cycle which was successful due to the selection of an egg that formed a high-quality embryo

Illumina are buying BlueGnome to push into the IVF market. Illumina wants to be a big diagnostics player and recently released test kits for Cancer and other diseases. IVF and fertility is another obvious clinical step.

BlueGnome also sell a library of 26,000 BlueFish fluorescent in-situ hybridisation probes. Probes are labelled by BlueGnome and sent directyl to users for running in their own labs removing the need to produce probes. FISH can be an important tool in cancer diagnostics as well, something Illumina can't have failed to spot.

You can find out more by watching BlueGnome TV.

Hey, I thought Agilent made BlueGnome chips: Yup, they do (did?) According to a March GenomeWeb report Agilent manufactures all BlueGnome's oligo-arrays. Although the CytoCHIP and 24sure are made internally with BAC probes and spotting. Exactly what the deal meams for Agilent is difficult to say but it is another blow from Illumina. Agilent and Illumina used to have cosy co-marketing deals but are more often chasing at each others heels over technologies related to sample prep.

The AGBT lottery opens soon

The 2013 AGBT meeting runs from the 20th to the 23rd of February and, as always, is in sunny Marco Island, Florida. This annual meeting is the one to watch for announcements around new genomics technologies. Last years hot topic was Oxford Nanopore’s announcement of MinION (although no-one appears to have head anything since). What will be the number one story this year? Well the 2013 meeting has an increased focus on clinical applications of NGS, so perhaps GnuBio will wow everyone with their technology, or perhaps someone will speak about exciting methods for analysis of circulating tumour DNA? ;-)

Buy your lottery ticket soon:  

Registration opens on October 1st and this year all applications will go into a clearing system to allocate the 1000 or so available spaces. One of the more common gripes from users has been the registration process. AGBT in 2011 & 2012 sold out in a matter of hours, the scientific equivalent of a a Stone Roses concert! This is not the conference organisers fault but rather shows how important the meeting is to users of NGS.

It perhaps also has something to do with being in Florida in February. Sun, sea and seqs anyone?

The organisers are going to limit the number of attendees from individual labs and institutes so expect arguments at Broad, Sanger and WashU in December. Registration includes the meeting, hotel costs and meals and the legendary AGBT party. Last year one post-doc was there in a home-made Storm Trooper costume that cost over $1000. I’m not sure what the theme is this year but you can expect more nerdy fun.

A list of things to do at AGBT:

•  Enter the competitions: last year I won an iPad2. I entered almost every competition going and there were over 20 big prizes on offer including iPads, Apple TV’s, etc. The odds are pretty good with around 1000 attendees and most not bothering to enter. 
• Attend the parties: last years parties were as good as usual. Lots of choice on offer and Agilents video wall was fantastic. A big headache with AGBT is the entertainment on offer, I am supposed to be working the next day for goodness sake! Oh well you can always sleep on the plane home. 
• Catch up in the bar: AGBT is a great place to catch up with colleagues you have not seen for ages. Everyone is in party mood and usually excited by the pre-AGBT announcements. Lots of ideas get floated around and who knows your next big collaboration might start over a beer.
  
• Don’t forget your bag: everyone at AGBT ends up with the same bag as they give out an exceptionally good one compared to most conferences (now I know where my $2400 registration fee goes). Wirte your name in your notebook or use a different bag for carrying round the meeting as it is too easy to get yours mixed up and lose all your notes. 
• Follow the conference on-line: MassGenomics, EdgeBio, Core-Genomics, Jonathan Eisen's blog, Omically Speaking, Pathogens Genes and Genomes all had coverage if you could not make it or get in and dont forget CrapBio, there #notAGBT announcement was the laugh of AGBT. Twitter will again have the #AGBT to follow. 
• Don’t forget the suncream: It was hot last year and swimming in the sea was certainly the order of the day. It being February I did not even think to bring suncream but fortunately I think Nanostring gave out little pumps of SPF50. Who wants to go home with sunburn! 
• Blogging and tweeting: I tried blogging through the whole meeting last year and am not sure I’ll try again. It is hard work making quickly scribbled notes understandable to a wider audience and it took too much time. Time I could have spent in the bar! There are very clear policies on blogging or tweeting at this years meeting although to go by last year most speakers are happy to get the coverage. I saw a large spike in readers of my blog and Core-Genomics received over 10,000 page views in February. Thanks to everyone that commented.

Changes to AGBT: One really annoying problem with AGBT (and all parallel session meetings) is the inevitable overrun of talks, stronger chairing and timekeeping would really help in moving between sessions to catch all the talks you want to hear.

There were hundreds of posters last year; far too many even to walk past. It would be great if these were all online, downloadable and searchable a few weeks beforehand so I could arrange to speak with the person presenting the poster with specific questions. It would also help if they were collected together under different themes using tags of some kind.

I’d like to see more people coming from the developing world this year. NGS is not just for the rich (although it helps to have a few $Million).

I’d also like to have a session for core facility labs, perhaps run by ABRF. Talking to other core labs is always useful, even if it just helps to reassure me that I am not the only person facing particular issues.



See you there (I hope).

PS: If anyone wants to share a room at next years meeting feel free to leave a comment. Double occupancy saves my budget a ton of cash and as my lab is funded by public donations I don’t see any reason not to share. Please don’t “apply” if you snore!

HuID-seq blog

There has been lots of recent activity around using NGS gene resequencing in the clinic. Although clinical DNA sequencing has been an important tool for several decades the explosion in NGS methods for amplicon resequencing has made it feasible for just about any lab to do. Previous posts on this blog have discussed NGS amplicon methods and some of the tools needed to design amplicons.

It is not easy to say clear which technologie(s) will dominate in the clinical space, nor whether small targeted panels will be preferred over more comprehensive and larger panels, medical exomes or even whole genomes. But it does seem pretty clear that amplicon-sequencing is going to be a very important clinical tool.

Why is patient ID important: As we are more easily able to sequence not just multiple genes but also multiple patients in s single NGS run it becomes very important to make sure results are not assigned to the wrong patient. Clinical molecular labs spend a huge amount of time and effort on making sure results don’t get mixed up, but I thought the tests themselves could be improved to determine which patient results came from at the same time as the clinical results are being generated. Just add a large enough number of SNP loci to allow patient identification by comparison to a simple blood-based genotyping assay.

SNP-seq for patient ID: I have been discussing using additional content in amplicon (and other) tests for a year or so, but have never found the time to get in the lab and demonstrate the idea. I asked our Tech-Transfer people about it and they said whilst it was a nice idea there was little that could be protected from an IP perspective. As I am not going to get time to work on it, and as it can’t be protected easily I am hoping this blog will help stimulate discussion and someone will take the idea on board for their research. I call the method HuID-seq.

Comparison to STR profiling: There is already a gold-standard for Human Identification in the STR profiling used in forensic applications. Unfortunately the tests cannot be simply added to an NGS assay. What we need is a level of discrimination so that results cannot be sent to the wrong patient, in theory the HuID-seq could be set at a level significantly lower than forensic STRs. Today 13 STR loci are used in the United States Combined DNA Index System (CODIS) forensic kits. SNPs have lower resolving power and more are likely to be needed but before I get onto that a recent paper deserves a mention. In Biotechniques Bornman et al published an NGS based method that reproduces STR data very nicely (see Short-read, high-throughput sequencing technology for STR genotyping). Potentially this could be added to current tests but I prefer SNPs for a number of reasons.

Why SNPs, which ones and how many: SNPs are a good choice because they are easily assayed by PCR amplification of by in-solution capture methods. SNPs are also already assayed by current NGS methods and perhaps most importantly if coding SNPs are used then RNA-seq data can also be used for HuID-seq. This will be important if array-based gene expression signatures are ported over to NGS. It also means that the HuID-seq method could be used to very good effect in research projects whatever the source of data. It is often important to quality control large experimental datasets to remove duplicate samples or wrongly assigned samples. In the supplementary information for the Nature paper The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups, the authors presented a novel eQTL-like approach to check that the same patients samples were used for whole genome gene-expression and SNP genotyping arrays. They termed the method BeadArray Diagnostic for Genotype and Expression Relationships (BADGER).

Which ones: There were around 250,000 SNPs present on both Affymetrix and Illumina arrays. 2000 could be used for identification, 100 for ethnicity and 200 for gender. All SNPs would allow mapping of samples onto other SNP or sequence based data, e.g. SNP arrays, ChIP-seq, RNA-seq and exomes or genomes. We looked carefully at which of these SNPs might be used in a HuID-seq method and came up with some requirements.

• MAF 0.5 (0.4-0.6)
• Present on most widely used genotyping arrays.
• Coding SNPs
• Ability to predict identity
• Ability to predict ethnicity
• Ability to predict gender

How many: The number of SNPs is going to be higher than the STR loci, but it need not be very high and the “real-estate” taken up by such a HuID-seq panel need not be large in comparison to an NGS clinical test. SNPs need to be present very broadly in the population to be of any use, ideally using SNPs with MAF0.5 gives any individual a 50:50 chance of being homozygous for one allele or heterozygous. If this 50:50 chance holds true for all SNPs, and if the assay used is perfect then (i.e. no errors in genotyping) then just 4 SNPs give a >90% chance of uniquely identifying an individual. Increasing this to 8 results in a 99.5% and 20 SNPs gives a 99.9999% chance. Choosing a set of SNPS such that there is at least one per chromosome arm leads to a set of about 48. The final number of SNPs used could be determined by the requirement for unique identification, cost or complexity of PCR.

Community cohesion: It makes sense that if this approach is going to be used then everyone should use the same set of SNPs. The best way to do this would be to get people from different clinical and research backgrounds in a room to discuss the why’s and wherefores’ of different SNPs and to recommend a set to the community. If Life Tech, Illumina, Roche, or others get marketing too early on then we are likely to see multiple standards. This is exactly what we have in STR kits with the US and Europe using different sets of loci.

There are already some SNPs being used to QC data by the Broad but the GATK pages have been updated and the older page is now a dead link! LifeTech are launching an 8 SNP Ampliseq based Sample ID kit giving a resolution of 1:5000, you spike the SNP targeting reagents into your assay and go. Illumina are aslo collaborating on forensic products, Dr. Bruce Budowle at the Institute of Applied Genetics, University of North Texas discusses forensic NGS applications on You Tube with the IlluminaInc channel.

It probably has not escaped your notice that a HuID-seq panel could be used for cell line authentication as well. This is something that often gets attention but is easily forgotten by PhD students and post-docs until too late. Making a test part of their ChIP-seq or RNA-seq experiment and comparing back to a reference database would be a simple ad-on.

My (almost) patent removing the need for users to quantitate before clustering

One of the things I most like about my job is being allowed to think. NGS has provided a very fertile ground for innovation and lots of novel ideas have become products we use regularly today. It is sometimes to easy to forget that any of us can have ideas that are just as good. The missing bit is usually the drive to try and commercialise or patent the idea.

I've tried several times with many ideas but have not been successful yet. In this post I will describe one idea I was sure was going to be the winning one, although in the end someone else had beaten me to it (gues who).

The problem: Illumina NGS systems require very careful quantification of the final sequencing library to generate cluster densities that give the highest yield from a flowcell. Getting QT wrong means over-or under clustering both of which lead to a drop in yield. And over-clustering can mean zero yield or lower quality data. Illumina recommends qPCR (although the method suggested has a few flaws) and many labs have had success wth qPCR, Bioanalyser, QuBit and other QT methods.

Standard Illumina clustering

But wouldn't it be nice if we could just put an unquantified library into a flowcell and always get the correct cluster density? I thought so and came up with a method to do just that.

The idea: basically I wanted to make sure only a certain number of discrete library molecules could hybridise to the flowcell surface for the initial hyb, but allow clustering to proceed normally afterwards. There are currently two oligos in the lawn on the flowcell surface, complementary to the ends of the Illumina adapters. I proposed adding a third “cluster seeding” oligo at a lower concentration. This oligo would have the same sequence as the current flowcell oligos but would be longer and include slightly more of the adapter sequence. Flowcells would be created with a similar lawn of primers such that the seed oligo is present at a spacing consistent with maximal yield and cluster requirements. Additionally the lawn oligos would be blocked by short complementary primers leaving only the 5’ end of the seed oligo unblocked for hybridization with library molecules. A covalently attached library molecule would be produced by polymerase extension. The first round of denaturation would remove the original library molecule and all blocking oligos. Bridge amplification would then proceed as normal.

This method would hopefully allow high, low and/or variable concentration of libraries to be entered into the flowcell without quantification as only a certain fraction would be able to hybridise. The yield of the flowcells would be significantly less variable.

This would remove the need for Illumina customers (you and I) to perform quantification and significantly increase yields from runs.

Seed-oligo clustering idea

Who got there first: Unfortunately when searching through patents to see if anyone else had a similar idea we found this patent US20090226975 by Illumina. They describe a similar method with a clever use of a hairpin oligo so the blocking is removed enzymatically. Fundamentally the same outcome is achievable and quantification can be considered a thing of the past.

US20090226975 figure 3

Why are we all running lots of qPCR: I don't understand why this has not made it out into the current generation of kits. All the users I talk to agree that removing quantification would be a good thing. Even if it is done perfectly it still takes an hour to do so time can be saved.

Come on Illumina, where is it?

Hopefully reading this has spurred you on to move ahead with the good idea you had a couple of months ago and left on the "back burner". I'll follow this post with another one about a completely different idea that is about to be realised by Life Technologies instead!

Ho hum, spending my millions will just have to wait!

TruSight blog

Yesterday Illumina released their clinical research NGS kits called TruSight.
TruSight: There are currently five kits in development, Cancer, Autism, Cardiomyopathy, Inherited disease and Human Gene Mutation Database Exome. Custom kits are sure to follow. Kits are designed to run on MiSeq ad the lab workflow is based on Exome capture protocols recently released as a new product from Illumina.

Combining Nextera with TruSeq (or any other) capture was a smart move (see a previous post on this blog). The simplicity of the library prep much better fits labs than the more complex standard adapter-ligation protocol. In the tests we did in my lab when we first tried the kits we found there was enough library from a 50ng input to sequence a small capture kit (TruSight for instance) and still have enough left over to capture an exome or possibly even sequence the whole genome.

GeneSight: Illumina also announced a partnership with Patners Healthcare on the GeneSight software. Aiming to make this the tool clinical researchers, medical geneticists and pathologists use for analysis. GeneSight is ready for MiSeq data, BaseSpace and the iPad MyGenome app, and is already FDA registered. It should be possible to provide a workflow from sample prep, through sequencing all the way to final analysis and interpretation. The press release mentions that geneSight has been used for over 24000 tests. If this takes off then expect to see graphs similar to ones showing how NGS yield has increased, but for the number of patients reported on by the MiSeq TruSight combo!

Geneinsight has been around since 2005 (this is a good paper describing it) and aims to help with data analysis is several ways. It acts as a repository for data in the form of case histories and variant information. It facilitates clinical reporting. And it will update clinicians as new variants become clinically relevant. This last feature is likely to cause some headaches as patients may need to be told their prognosis has changed based on new information. The GeneInsight network also allows labs to share results and new findings increasing the ease with which variants might be understood to be clinically relevant.

Most of us are going to have to wait though. As is all too often the case with launches like this we have the information on how exciting this is going to be but no release date. A few pilot sites and Illumina’s own CLIA labs will be the first to access the package.

Competition within Illumina: TruSight will go hand-in-glove or head-to-head with the TheraSeq product Illumina recently made a pre-release announcement about. TheraSeq uses the TruSeq custom amplicon approach to target small numbers of clinically relevant genes, see a TheraSeq overview for more details. Currently in development are kits for Non-small cell lung cancer, Advanced metastatic disease and Gastrointestinal stromal tumor. It will work from FFPE tissue and again is likely to run on MiSeq.

If you want to steer Illumina's clinical developments then why not take the survey on the TheraSeq page? One of the interesting questions relates to the size of test panels asking if responders prefer small or large panels, and whether these should be dictated by current standard-of-care or expand beyond it. They also ask for opinions on returning variants of unknown significance. Both hot topics in diagnostic panel discussions.

Also on the TheraSeq page is a link to an overview of Cancer research papers that used Illumina technology.

Summary: Illumina, and all the other NGS companies, know how important clinical is going to be in their future growth. This is unlikely to be just through instrument, consumable and test kit sales. Service provision could be an important revenue generator and this is likely to put Illumina into direct competition with the people they are selling to today. Although according to GenomeWeb Matt Posard said that “Illumina does not intend to offer its TruSight assays as a diagnostic service, because that would compete directly with its customers.

Radiogenomics is coming

I posted last week about the emerging field of Immunogenomics. Today I’ve taken a brief look at what is happening in Radiogenomics. Whilst this field is not using NGS in such a comprehensive way I think it can only be a matter of time before it ramps up.

Radiotherapy is an important tool in treating cancer and the impact of genomics on the field was recognised by researchers in Cambridge and Manchester in 2004. Those researchers started the Radiogenomics: Assessment of Polymorphisms for Predicting the Effects of Radiotherapy study (RAPPER) and also helped found the International Radiogenomics Consortium.

The ultimate aim of this consortium is to individualise radiation dose prescription for patients maximising the impact on the tumour whilst minimising normal tissue damage for the patient. They aim to find genomic variants that can help predict how patients will respond to radiotherapy and allow tailoring of treatment. This is somewhat similar to pharmacogenomics approaches used for drugs like Warfarin where SNP genotyping can help establish the correct dose for individual patients. The consortium should make it easier to collect samples for genomic studies and also spur development of methods for radiogenomic research.

The consortium is likely to also learn a lot more about the biology of radiation-induced tissue and DNA damage. Whilst understanding how individuals may respond to radiotherapy is a primary goal, hopefully a better understanding of biology may lead to a list of genes that might be mutated in tumours making them more susceptible to radiotherapy as well.

The Radiogenomics Consotium conducted a GWAS in radiotherapy patients (Independent validation of genes and polymorphisms reported to be associated with radiation toxicity: a prospective analysis study. Lancet Oncol. 2012) to address concerns over how underpowered previous research on late side-effects had been. Late side-effects can have serious impacts on patients and their treatment. This prospective study genotyped 92 SNPs (selected from previous studies) in 1600 breast and prostate cancer patients using the Fluidigm 96.96 Dynamic Arrays. None of the SNPs previously reported to have a significant associations with radiation sensitivity were confirmed. The consortium suggested that the previous associations were “dominated by false-positive associations due to small sample sizes, multiple testing, and the absence of rigorous independent validation attempts in the original studies”.

As the costs of sequencing continue to fall and as associations are found it is likely that NGS will become a more important tool for the consortium. Longitudinal studies of cancer patients can be incredibly revealing and comparison of cancer genome and normal genome with radiotherapy follow up data is likely to yield interesting results.

Immunogenomics is coming

The immune system is becoming easier to investigate as new methods based on nextgen sequencing are published. I am not an immunologist and the complexities of the immune system for me are stuck back in the days of my undergraduate training. And that was in the 90’s!

Nature and the HudsonAlpha Institute are hosting the first Immunogenomics conference next month bringing together scientists from many disciplines to learn about large-scale immune sequencing projects; epigenetics and the immune system and many other topics. Immunogenomics looks like it is going to make headlines next year.  

There have been several papers describing HLA typing (e.g. Gabreil 2009 & Bentley 2009 using 454 and more recently Wang 2011 using HiSeq) and many groups are working on using next-gen methods to replace older tests.
 
A new product from Sequenta is aiming to make this kind of analysis simple to do for any user. The Lymphosight platform uses a multiplex PCR to amplify the IgH, IgK, TRB, TRG, TRD immune cell receptor loci, allowing each T or B cell to be characterised and counted. Immune cell proliferation in response to disease and other studies might be far easier to carry out using this new kit. With 100’s of millions of reads coming from HiSeq, and eventually Proton, even fairly rare immune cells should be detectable in a high background. 

Lymphosight workflow from Sequenta website

The company discuss a test they ran where sequences associated with a B cell tumour were diluted into a normal background at 1:1,000,000. They got very reproducible, quantitative results and a useful dynamic range that compares well to flow cytometry methods currently being used. They expect Lymphosight to be useful in monitoring of minimal residual disease.

David Haussler Director, Centre for Biomolecular Science and Engineering at UCSC said 'We can read genomes from your immune cells. They adapt throughout your lifetime so they can protect you from diseases. Reading those genomes will be important, and you’re going to hear a lot about them next year.'

I recently visited TRON, a spin out from Mainz University Medical Center, where they are conducting translational research in the field of oncology and immunology. One of their aims is to take personalised immunogenomic markers and turn these into personalised Cancer vaccines. The head of TRON, Ugur Sahin just published a very interesting article in OncoImmunology where they describe using NGS to demonstrate a proof-of-concept for identification of immunogenic tumour mutations that are targetable by individualised vaccines. They analysed a melanoma cell line and found over 500 non-synonymous expressed somatic mutations, one third of which were immunogenic. From these they made long peptides of 27aa length and tested these for immune response. 11 of these immunogenic tumour-specific peptides effectively immunised mice against the Tumorigenic cell line (see figure 1 from their paper below).

I am sure Sequenta are hoping groups like these will be using Lymphosight to do perform their analysis of the Immune repertoire.

How to make Outlook "out-of-office" work for you

Apologies to readers who might have been hoping for some posts over the past few weeks but I have been offline whilst holidaying in France and Spain.

One of my "to-do's" before I left was to respond to the "change your password or you'll be locked out" email from our IT manager. This was one thing that got missed at the end of the frantic Friday afternoon before leaving and subsequently I could not log on, even through webmail.

However, this made for a lovely and uninterrupted holiday and I am sure made it easier to forget about work as there was little point to even try and get online. As a result though I have come back to over 800 emails.

This got me wondering about what I might be writing in my out of office message next time I go away. I think it should go something like this...

"I am currently out-of-the-office on holiday until the 1st of September and will not have any access to email whilst away. If your message really is important then please send it again on the 2nd of September as I will be deleting every email I receive between now and my return to work. Sorry for any inconvenience."

This would certainly make the task I now face much easier!

PS: Normal posting will resume shortly.