5:43 AM
Todd Kuehnl has been a developer for nearly 20 years and says he's tried "pretty much every language under the sun."
But it was only recently that Kuehnl discovered Go, a programming language unveiled by Google almost four years ago. Go is still a new kid on the block, but for Kuehnl, the conversion was quick. Now he says "Go is definitely by far my favorite programming language to work in." Kuehnl admitted he is "kind of a fanboy."
I'm no expert in programming, but I talked to Kuehnl because I was curious what might draw experienced coders to switch from proven languages to a brand new one (albeit one co-invented by the famous Ken Thompson, creator of Unix and the B programming language). Google itself runs some of its back-end systems on Go, no surprise for a company that designs its own servers and much of the software (right down to the operating systems) that its employees use. But why would non-Google engineers go with Go?
This is just a picture, but you can try Go within the browser at golang.org.
Go
Kuehnl is the head of engineering for Beachfront Media, which recently announced a new platform called Beachfront.iO that serves video ads to mobile devices and tablets. Kuehnl spent three months leading development of the platform, saying he chose Go in large part because of how it enables concurrency, the ability for computations to execute simultaneously and interact with each other while they're running.
Beachfront.iO "needs to service millions upon millions of requests a day, upwards of 5 to 10 thousand transactions per second and do this very reliably," Kuehnl said. "I was looking at different stacks and doing benchmarks to figure out how it would be most productive and most performant … and that's where I came across Go."
The code written in Go performs all the heavy lifting on the back-end, including load balancing and choosing which ads to serve up when and where.
"The issue for PHP and even Node.js is obviously you're trapped in a single-threaded situation and what I really wanted was to be able to do a lot of things concurrently," Kuehnl continued. "My options were to go with something like Java, where you have more memory overhead, or I could go with something like Go that was built from the ground up for concurrency and using very modern patterns."
Kuehnl said Go combines those modern concurrency patterns with the "static execution speed of C or C++" but with "a more compositional feel, a script kind of feel. … I started first with the idea of trying to pick the most high-performance modern language, but as I explored it more the beauty of the language presented itself."
Beachfront runs its video ad service on Ubuntu Linux servers hosted in the Amazon Elastic Compute Cloud. Go performs well even on cloud-based virtual machines (albeit large instances with 16GB of RAM), Kuehnl said. This allows Beachfront to minimize one of the pain points of advertising—lengthy delays that cause users to give up and move to a different website or app. "With Go, I'm able to have an ad up and running within 200 milliseconds, which is orders of magnitude faster than some of these legacy networks," he said. Go's compiler is so fast "it's almost like working in an interpreted language," he said.

He's not the only one

Google unveiled version 1.1 of Go earlier this month, promising a big speed boost. (Kuehnl said he's already using the new version.) Google, which is also developing a JavaScript alternative called Dart, declined an interview request from Ars about the future of Go. Google noted, however, that Go is used internally by the company for "core infrastructure services; cluster management; large-scale distributed processing, including some uses of MapReduce; [and in] tools for software development."
Google pointed out that Go is the 24th most popular language on GitHub, and that interest in Go is growing rapidly on Google Trends. The language is being used by a variety of companies including BBC, Bit.ly, Canonical, CloudFlare, Disqus, Mozilla, and Tumblr.
At least one project using Go did end in failure. That was a Kickstarter-funded game called "Haunts: The Manse Macabre." The developer blamed much of the trouble on Go itself, but his descriptions seem to indicate the problems were mostly self-inflicted through poor planning and version control. The developer cited "the shifting code base of the Go programming language" and the fact that the language is "new and not well supported." But he also noted that the code his team developed is "buggy and incomplete" and that the original programmer on the project took on a new job and "has not responded to additional requests for aid or insight into solving" the fundamental problem that the game code can't be compiled.
Go has worked out well for numerous other users. A developer named Dotan Nahum recently used Go to build a remote management system for the Raspberry Pi called Ground Control.
Nahum told me via e-mail that he chose Go for its "performance, concurrency, and simplicity." Ground Control required a Web server and monitoring daemon, and his options "were to either go with C and write a pure C-based solution, or use something I knew was the closest thing to C in concept and offered great performance (in its recent version): Go."
Nahum wasn't sure how Go would work on the Pi's ARM architecture, but it turned out the Pi "took everything I threw at it without any problems."
Nahum further stated:
Being polyglot for a long time now, I use the right tool for the job. Instead of being stuck fitting square pegs into round holes (which sadly many programmers do), I keep exploring more and more platforms and languages in order to gain high expertise in those and understand what situations call for the right platform and language.
In my opinion Go gives you great performance—which is easy to verify yourself and is more or less a fact; but what's more arguable is that I think it gives you great concurrency. More precisely I think it gives you great concurrency, and simple concurrency, for the amount of effort you put in.
I see Go as simple in general, too. It doesn't try claim it is a superstar like other languages have in the past, it has simple language constructs (in my opinion), and the ecosystem is very humane—from the documentation to the developers' general approach. I think you can attribute those properties to languages like Ruby and Clojure too—I see Go as a distant relative.
Go still has some growing up to do, though—no surprise given how new it is. Nahum said alternatives like Ruby and Node.js have more complete ecosystems and that Go's packaging and dependencies could stand some improvement. "If you make a few searches, you'll see some confusion there and several alternate solutions for what exists right now to specify your dependencies," Nahum wrote. "Some say that it's just a matter of people not getting it right, and some say that it shouldn't work the way it does right now. I'm not so fond of the way it works right now, but I can also make do for now—it could have been much worse (I don't want to say which language I'm talking about :)."
Kuehnl noted that Go is probably not yet the best choice for creating desktop apps with graphical user interfaces, or for gaming. "I think Go fits best for what it's really being used for in Google, which is back-end services, back-end processes, data analysis, things like that," he said.
Kuehnl watched some video lectures and read documentation about Go when he was learning the language, and said it didn't take him to figure out how best to use it.
"There's sort of a C kind of feel to it but it's definitely different, it's got some influences from Haskell and some other languages in regard to some of the concurrency aspects," he said. "I found Go to be very natural really within a few days. What I love about Go is it gives all these powerful concurrency patterns and you can get really creative with it. At this point I don't have to think about it. I just think about the problem I'm trying to solve and there's a lot of really elegant ways to go about it with Go."


5:41 AM
















Modesto Orozco, IRB Barcelona (Barcelona, Spain) and Charles A. Laughton, Nottingham University (Nottingham, UK)

Barcelona, Spain. 4-6 November, 2013

Recent algorithm improvements and the development of new computer platforms (supercomputers, GPUs, specific purpose computers) is leading to a revolution in the field of molecular dynamics simulations of biomolecules. The field evolves extremely fast, millisecond scale simulations are possible, the dynamics of supramacromolecular systems containing millions of atoms is determined and projects exist to transform structural databases into dynamical databases. Clearly, molecular dynamics is moving structural biology to a different scenario.
The BioMed conference will be a forum to meet many of the most distinguished scientists in the field. We will not only learn on recent state of the art simulations, or on the latest theoretical or algorithmic improvements, but we will also discuss on the integration of molecular dynamics with experimentation, and on the challenge that the technique will face in a close future dominated by the omics research.

Registration deadline: September 15, 2013
There is no registration fee for this conference, but the number of participants is limited.
Participants are invited to submit abstracts, a number of which will be selected for short talk and poster presentations. Abstracts should include a title, authors, affiliations, summary (max 250 words) and references.

5:39 AM
University of Cincinnati (UC) researchers played a pivotal role in two recent cancer studies in which he used bioinformatics—specifically, public domain genomics data—to help identify a tumor suppressor gene’s role in human cancers. The approach took the work out of animal models and moved it into computer analysis.
Bioinformatics is a relatively new field of science which incorporates biology, statistics, computer science and information technology. By using large data sets and new statistical models, scientists can make discoveries or learn new insights into human disease.

Mario Medvedovic, PhD, an associate professor in the department of environmental health at the UC College of Medicine, and Jing Chen, PhD, research scientist in Medvedovic’s group, co-authored two recent National Institutes of Health-supported studies appearing in Cell and Proceedings of the National Academy of Sciences. Both studies relied on bioinformatics analysis of genomics data.

"An interesting aspect from our angle is that in both papers we used public domain genomics data to connect experimental genomics data from in vivo and in vitro models to human diseases,” says Medvedovic, whose lab has been actively gathering and processing public domain genomics data and developing bioinformatics to analyze these data. It has created web servers where anyone can go and mine these data. (http://GenomicsPortals.org and http://LincsGenomics.org)

"Re-use of public domain genomics datasets is a pretty hot topic locally and nationally.” (The U.S. Supreme Court recently heard arguments for and against patenting of human genome data. A decision is expected in June.)

Tumor Suppressors and Cancer
Much of our understanding of cancer comes from research that uncovers the molecular interactions underlying this disease. The discovery of tumor suppressor genes that control cell division and growth provides great insight into how cancer develops.

Tumor suppressors are viewed as protective genes that prevent the uncontrollable division of cells, a hallmark of cancer development. A disruption in the function of these genes can catapult cancer progression. Protein Kinase C zeta (PKC zeta) is a tumor suppressor known to play a role in different human cancers. A mutated form of PKC zeta found in people is associated with tumor development.

The exact mechanism by which PKC zeta (or lack of) affects the progression of cancer is not well understood. Recently, a team of researchers led by scientists at Sanford-Burnham Medical Research Institute identified PKC zeta’s role in prostate cancer and its mechanism of action. The same authors previously published a paper demonstrating PKC zeta to be a tumor suppressor in human and mouse intestinal cancer.

This Research: Before Bioinformatics
Prior to utilizing public domain genomics data and bioinformatics, the researchers used a mouse model deficient in tumor suppressor PTEN, which predisposes the mouse to cancer, to determine PKC zeta’s role in prostate cancer. Using this model, the researchers found that the loss of PKC zeta resulted in prostate cancer. These results directed them to establish PKC zeta’s role in human prostate cancer in the context of PTEN deficiency.

They also analyzed protein expression of PTEN and PKC zeta in human prostate cancer tissue samples and saw that there was a positive correlation between the expression of two tumor suppressors.
Using Public Domain Datasets

The researchers wanted to further assess the PTEN-dependent role of PKC zeta in normal prostate tissues, primary cancer and metastatic primary cancer tissue. With the help of Medvedovic and Chen, the team was able to establish PKC zeta’s role in human prostate cancer through bioinformatics analysis.

By mining their collection of genomics datasets, Medvedovic and Chen determined the appropriate human prostate cancer dataset to use, which provided information about gene-level changes in prostate cancer. The results of this analysis showed that reduction in expression levels of PKC zeta in metastatic cancers is dependent on a decrease in PTEN expression levels. This supported previous findings on PKC zeta’s role in colorectal cancer, and from there, the researchers were able to explore even deeper the cellular mechanism of PKC zeta in prostate cancer.

To investigate the molecular mechanisms by which PKC zeta restrains prostate cancer, researchers performed a genome-wide transcriptome analysis and identified genes that are differentially expressed between cells without PKC zeta and cells with PKC zeta. Expression levels of these genes were then used to cluster samples in the human prostate cancer datasets.

The researchers found similarities among the altered genes in cells and tumors without PKC zeta when compared with normal tissue. Bioinformatics analysis of the altered genes showed that they are involved in proliferation, growth, movement and cell death, suggesting that, when present, PKC zeta plays a role in preventing progression of prostate cancer, invasion and metastasis.

Further analysis of these differentially expressed gene signatures led to the identification of another gene
(c-Myc) as a relevant target of PKC zeta. Levels of c-Myc increase when PKC zeta is absent, and this was confirmed in the prostates of the PTEN mouse model deficient in PKC zeta. These data suggest that the loss of PKC zeta results in increased levels of c-Myc, subsequently affecting cell proliferation and growth.

Medvedovic and Chen then used a public domain ChIP-seq dataset and their own bioinformatics technique for analysis of transcription factor DNA-binding patterns to identify c-Myc targets. They found that genes upregulated in cells without PKC zeta tend to be targets of c-Myc, further underscoring the idea that PKC zeta negatively regulates c-Myc to prevent cancer. The researchers hypothesized that PKC zeta not only repressed the gene levels of c-Myc but also repressed by directly acting on c-Myc. 

The application of bioinformatics tools allowed the researchers to:

• Correlate the PTEN-dependent role of PKC zeta to human prostate cancer.
• Identify and categorize genes regulated by PKC zeta.
• Identify the clinically relevant target of PKC zeta, c-Myc.
• Provide supporting evidence that PKC zeta regulates cell division and growth by negatively regulating c-Myc, thereby preventing the development of cancer.

"Prostate cancer is the most common malignancy among men in Western countries,” the authors write. "Our observations that PKC zeta is a tumor suppressor in this type of neoplasia, and that it acts by repressing c-Myc expression and function, are likely to be highly relevant in the design of new therapeutic approaches, which are sorely needed.”

Adds Medvedovic: "The vast amount of diverse public domain genomics datasets provide a tremendous opportunity to test and postulate new hypotheses by simply re-analyzing other people’s data. Such data can also be used to better interpret results of laboratory experiments and increase their impact by making a connection with human disease.”

Medvedovic is a member of the UC Cancer Institute. Both Medvedovic and Chen work with UC’s Center for Clinical and Translational Science and Training (CCTST) to assist investigators with bioinformatics analyses. The CCTST is the academic home of UC’s institutional Clinical and Translational Science Award (CTSA) from the National Institutes of Health.

Jessica Dade is a fourth-year graduate student in molecular genetics, biochemistry and microbiology with an interest in science writing. She participated in UC’s summer research program in 2008 and began her graduate work at the university the following year. She is working in a lab at the Cincinnati Department of Veterans Affairs under the direction of George Deepe, MD, and George Smulian, MD, both of UC’s infectious diseases division.
Media Contact:     AHC Public Relations, (513) 558-4553

5:34 AM

Law's Laws

A series of observations on Genetic Analysis algorithms and experiments

Law's First Law

"The first step in developing a new genetic analysis algorithm is to decide how to make the input data file format different from all pre-existing analysis data file formats."
A prime exemplar of this Law is the use of different codes to signify the sex of animals. For example, crimap uses '0' to represent female and '1' to represent male. The algorithm designed by Keightly et al. uses the same codes to mean the opposite sexes.The Knott & Haley QTL analysis algorithm uses codes '1' and '2'. The list goes on.

Law's Second Law

"Error messages should never be provided"
corollary...
"If error messages are provided, they should be utterly cryptic so as to convey as little information as possible to the end user"
Do you understand crimap's error messages? I thought not.

Law's Third Law

"The number of unique identifiers assigned to an individual is never less than the number of Institutions involved in the study"
... and is frequently many, many more.

Law's Fourth Law

"All scientists agree that sharing data is good and are more than happy to share everyone else's"

5:32 AM



 
 
 
 
Reference ID:
795638
Company:
St. Jude
Position:
Bioinformatics Associate Research Scientist (two openings), Computational
Requirements:
PhD in Molecular Biology, Biochemistry, Computer Science, Statistics, Mathematics, Bioinformatics, or related field required. PhD which must include research related to bioinformatics (such as analysis of sequence data, microarrays, SNPs, image data, proteomics data, or biological pathways; development of algorithms, statistical methods, or scientific software); OR If PhD with no bioinformatics research, then two (2) years of pre-or postdoctoral experience in Computational Biology or Bioinformatics research is required.
Experience with programming languages such as Perl, C, or Java required.
LICENSURE REQUIREMENTS:
None
OTHER CREDENTIAL REQUIREMENTS:
None
Published:
5/22/2013 11:41:06 AM
Location:
Memphis,TN,US

St. Jude Children's Research Hospital, founded by the late entertainer Danny Thomas, is one of the world's premier centers for the research and treatment of pediatric cancer and other catastrophic childhood diseases. St. Jude is the first and only pediatric cancer center to be designated as a Comprehensive Cancer Center by the National Cancer Institute. Children from all 50 states and from around the world have come through the doors of St. Jude for treatment, and thousands more around the world have benefited from the research conducted at St. Jude - research that is shared freely with the global medical community. St. Jude is the only pediatric cancer research center where families never pay for treatments that are not covered by insurance. No child is denied treatment because of a family's inability to pay.

Job Description:
The Computational Biology (CompBio) Department focuses on the development and application of innovative approaches for analyzing high-throughput; multi-dimensional genomic and epigenetic data generated from basic and clinical research groups studying pediatric cancer, gene therapy and infectious disease. The Department has a well-established track record in developing state-of-art computational methods for analyzing next-generation sequencing (NGS) data with high impact publications in the journals of Nature, Nature Genetics, Nature Methods, JAMA and Cancer Cell. We are looking for highly motivated and talented bioinformatics scientists who are interested in working on a large-scale clinical sequencing project with responsibilities for in-depth analysis of whole-genome, exome and RNASeq data of pediatric cancer patients. CompBio provides a highly collaborative teamwork environment, access to state-of-art computational infrastructure and deep experience in analyzing, managing, visualizing and delivering data and results generated from NGS technology.

The Bioinformatics Associate Research Scientist in the Computational Biology Department is expected to participate in data analysis, data visualization, statistical analysis, experimental design, and database development. Provides bioinformatics analysis for interdepartmental investigators and communicate and discuss with investigators on analytical process and results. Participates in the Computational Biology Department's independent research. Assists in preparing and submitting manuscripts for publication. Contributes ideas to automate or improve existing analysis methods. Assists with establishing and documenting protocols or best practices for common research tasks. Ensures that efficient and prompt help is provided to the SJCRH investigators. (MJP)




5:25 AM
Job: NERC Bioinformatics fellowships at UCL
From the Evolution Directory (EvolDir) via Twitter.

Dear all,

UCL's Research Department of Genetics, Evolution and Environment invites expressions of interest from potential applicants to NERC's Independent Research Fellowships in Bioinformatics.

NERC has recently launched a research programme in "Mathematics and Informatics for Environmental 'Omic Data Synthesis" and funds 5-year fellowships for early-career scientists wishing to establish independent research groups. More information on the scheme and the background of the programme can be found on the NERC website (http://www.nerc.ac.uk/research/programmes/omics/events/ao-bioinformaticsfellowships.asp).

If you have the appropriate expertise and would like to apply for a fellowship hosted in our department, please get in touch with Max Reuter (m.reuter@ucl.ac.uk). We will support selected candidates through all stages of their application.

Our department fosters young talent and provides a stimulating and multi-discilinary research environment. It has strengths in evolutionary and statistical genetics, genomics, evo-devo and environmental and biodiversity research. For more information about our research, please visit the department's website (http://www.ucl.ac.uk/gee/) as well as those of departmental sub-nuits including the UCL Genetics Institute (http://www.ucl.ac.uk/ugi/) and the Centre for Biodiversity and Environment Research (http://www.ucl.ac.uk/cber/).

Best regards, Max


PS: Sorry for posting this twice, but I wanted to make sure it went to people looking at 'Job' and 'Postdoc' messages



___________________________________________________________
Max Reuter


Research Department of Genetics, Evolution and Environment
Faculty of Life Sciences
University College London
Darwin Building
Gower Street
London WC1E 6BT, UK

Phone: +44-20-76792201 (internal 32201)

Lab: http://www.homepages.ucl.ac.uk/~ucbtmre/Labsite/
Department: http://www.ucl.ac.uk/gee
Centre for Ecology and Evolution: http://www.ceevol.org.uk
___________________________________________________________

5:22 AM















This year’s symposium will be jointly presented by the Kansas City Area Life Sciences Institute and Frontiers: The Heartland Institute for Clinical and Translational Research. The symposium will be open to the public and feature an outstanding group of national and regional speakers. We look forward to your attendance!

View Agenda    Register Now

Date & Time

June 20, 2013Light Breakfast 7:30 – 8:00 am
Program 8:00 am – 5:30 pm

Location

Kansas City University of Medicine and Biosciences
Weaver Auditorium
1750 Independence Avenue
Kansas City, MO 64106
Directions
KCUMB
Hosted by Kansas City University of Medicine and Biosciences

5:20 AM

 

 

 

 

 

 

PLATO, an Alternative to PLINK

Since the near beginning of genome-wide association studies, the PLINK software package (developed by Shaun Purcell’s group at the Broad Institute and MGH) has been the standard for manipulating the large-scale data produced by these studies.  Over the course of its development, numerous features and options were added to enhance its capabilities, but it is best known for the core functionality of performing quality control and standard association tests. 

Nearly 10 years ago (around the time PLINK was just getting started), the CHGR Computational Genomics Core (CGC) at Vanderbilt University started work on a similar framework for implementing genotype QC and association tests.  This project, called PLATO, has stayed active primarily to provide functionality and control that (for one reason or another) is unavailable in PLINK.  We have found it especially useful for processing ADME and DMET panel data – it supports QC and association tests of multi-allelic variants.    

PLATO runs via command line interface, but accepts a batch file that allows users to specify an order of operations for QC filtering steps.  When running multiple QC steps in a single run of PLINK, the order of application is hard-coded and not well documented.  As a result, users wanting this level of control must run a sequence of PLINK commands, generating new data files at each step leading to longer compute times and disk usage.  PLATO also has a variety of data reformatting options for other genetic analysis programs, making it easy to run EIGENSTRAT, for example.

The detail of QC output from each of the filtering steps is much greater in PLATO, allowing output per group (founders only, parents only, etc), and giving more details on why samples fail sex checks, Hardy-Weinberg checks, and Mendelian inconsistencies to facilitate deeper investigation of these errors.  And with family data, disabling samples due to poor genotype quality retains pedigree information useful for phasing and transmission tests. Full documentation and download links can be found here (https://chgr.mc.vanderbilt.edu/plato).  Special thanks to Yuki Bradford in the CGC for her thoughts on this post.  

5:18 AM
Novus Explorer — your gateway to scientific research. This bioinformatics tool is designed to facilitate scientific exploration of related genes, diseases and pathways based on co-citations.  Type in your gene, disease or pathway of interest and press Enter to begin your exploration. Read the Novus Explorer How to Guide for more detailed instructions. If you are having problems with Flash please try clearing your browser cache.

5:17 AM
Bioinformatics outsourcing. Image source pharmatching.com
Never before has Research and Development (R&D) been witness to such turbulence hitting its ranks. Rising costs of research and tight budgets have led laboratories to look for new nesting grounds. Lack of skilled personnel and new government legislations have led to rising concerns about the future of Bioinformatics.
As the industry is looking at alternatives to revitalize and renovate, outsourcing has emerged as a potential savior for its revival.
A report that was published in December 2012 predicts that in 2018 bioinformatics will reach a market size of value USD 9.1 Billion. The market report “Bioinformatics Market – Global Industry Size, Market Share, Trends, Analysis and Forecast, 2012 – 2018” published by Transparency Market Research forecasts a double-digit growth that will be a record on its own. The bioinformatics platform sector is expected to contribute the highest revenue.
In 2012, the bioinformatics market was estimated to be worth USD 2.3 billion. The record growth during the period 2012—2018 is expected at 25.4% CAGR. This rapid rate of growth is attributed to the demands from related fields such as pharmaceutics, health care, medicine, diagnostic research, biotechnology and various other sectors linked to life sciences.

Bioinformatics Outsourcing offers an effective workaround

Integration of various services is one of the main elements that impel outsourcing in the bioinformatics sector. External data can be acquired and integrated into internal systems in an easy manner.
Companies based on Bioinformatics R&D are on the lookout for vendors that can carry out operations involving search, analysis, sharing, and customization of data. These types of services have to be backed up by consultancy and after-sale services.
Mauritius is one country that has realized the potential of bioinformatics outsourcing. Significant research is being conducted in this country in this field. Bioinformatics outsourcing is set to take a great leap in this country.
A large talent pool of graduates and scientists in various fields of life sciences adds to the inevitable growth of this industry in the region. The rich flora and fauna together with diverse species of plants with medicinal value aid in the spurt of genomics and proteomics research.
Companies specialized in Bioinformatics R&D can thus improve their future prospects by tapping the vast amount of resources that have remained hidden in other regions of the world. It is for them to take up this challenge and carry this industry forward into hitherto unknown avenues that translate into success and development on a global scale.
It is safe to assume bioinformatics outsourcing has a bright future.

5:15 AM
A semiconductor sequencing chip held by Jonathan Rothberg, founder of sequencing company Ion Torrent in Guilford, Connecticut — one of only two entrants for this year’s Archon Genomics X Prize.
MICHELLE MCLOUGHLIN/REUTERS
It was never meant to be a piece of cake — but neither was it meant to be a flop. Yet as the 31 May registration deadline looms for the Archon Genomics X Prize — a challenge to sequence 100 complete human genomes in 30 days at unparallelled accuracy and low cost — only two teams have entered.
The lacklustre showing is a testament to both the difficulty of the challenge and the maturation of the DNA-sequencing industry in the seven years since the prize was first conceived, genetics and innovation researchers say.
“The business has become bigger than the prize,” says Jonathan Rothberg, founder of the sequencing company Ion Torrent in Guilford, Connecticut, which was acquired in 2010 by Life Technologies in Carlsbad, California — which was, in turn, recently snapped up for US$13.6 billion by Thermo Fisher in Waltham, Massachusetts. Ion Torrent plans to compete, but other firms have apparently decided that they have little to gain.
Yet the goal of the prize — to drive down the cost of sequencing while improving its quality — matters just as much as it did in October 2006, when the X Prize Foundation, based in  Playa Vista, California, first announced the challenge, experts say. Although sequencing costs have fallen drastically (see ‘Plummeting costs’), that decline has plateaued recently.
Expand
SOURCE: NHGRI
The original rules called for teams to sequence 100 genomes in 10 days for less than $10,000 per genome. After none of the original eight competitors could meet the 10-day timeframe, the foundation spent two years revamping the challenge. The reconceived prize, launched in October 2011, extended the time to 30 days, tightened the cost to $1,000 per genome and specified that 100 genomes from centenarians, who may harbour life-extending genetic variants, must be sequenced (see L. Kedes and G. Campany Nature Genet. 43, 1055–1058; 2011). 
The new challenge aims at what the X Prize Foundation calls a “medical grade genome” — a sequence of all the nuclear DNA to 98% completeness and high accuracy, allowing only one error per million bases. To win the $10-million prize pot, teams must also find DNA insertions, deletions and rearrangements within genes and determine which parent each one came from.

Hitting all these goals in one go is hugely challenging. Market leader Illumina in San Diego, California, boasts a rate of false-positives (inaccurately flagging a DNA base as a variant from normal) of 0.25% and a rate of false negatives (missing a real variant) of 7.4%, well above the error rates allowed by the X-prize requirements. It is hard to do better because current technologies sequence the genome in short stretches that then have to be reassembled, introducing errors. New ways to sequence longer segments in one go are commercially available (see Nature 494,290–291; 2013), but they are slow and expensive. “To date, none of them would win the X prize at this scale,” says quantitative biologist Michael Schatz at Cold Spring Harbor Laboratory in New York.
Still, why have so few teams even deigned to try? Meeting the challenge would cost much more than the prize purse, but that has also been true of past contests that attracted dozens of entrants, such as the Ansari X Prize, which required teams to send passengers into space, and the US government’s Defense Advanced Research Projects Agency Grand Challenges, one of which catalysed the development of successful self-driving cars.
Part of the answer is that a genomics prize, unlike a rocket launch, isn’t easy to explain to the public (see 'Other challenges'). As a result it does not have the same publicity value, says Luciano Kay, a researcher at the Center for Nanotechnology in Society at the University of California, Santa Barbara. A competition for a self-driving car that can go 10 kilometres is more attractive than manipulation of matter or genes at tiny scales to accomplish a very scientific or technical feat, Kay says.

Other challenges

The X prize isn’t the last word on genomics competitions. Cheaper to solve than technology-based contests, bioinformatics challenges have proved hugely popular.
The Critical Assessment of Genome Interpretation experiment has seen the number of entrants rise every year since it was first held in 2010 (see go.nature.com/dfclt1). And a US$1-million purse offered by the US Defense Threat Reduction Agency has spurred thousands of researchers to try to identify individual organisms from mixed pools of sequenced DNA.
Thousands of participants also competed in a bioinformatics challenge in April to find hidden sequences in DNA data sets (run by the journal Genome Biology and curated by Michael Schatz of Cold Spring Harbor Laboratory in New York and James Taylor of Emory University in Atlanta, Georgia). It was organized to commemorate the sixtieth anniversary of the original research papers describing the structure of DNA.
Now Grant Campany, senior director for the Archon Genomics X Prize, says that the X Prize Foundation itself is considering laying out another challenge aimed at genome interpretation.
More
And the goal of the genomics X prize — to sequence whole genomes to medical grade rapidly and cheaply — may not be a top commercial priority at present. The business of genomics is already booming on the basis of less complete sequences, and Rothberg points out that scientists can only interpret the small fraction of the genome that codes for proteins (the ‘exome’). It is unclear what would be gained from an accurate sequence of the rest.
Still, whoever wins the prize earns the right to boast, which explains why Ion Torrent decided to compete. It also explains why Illumina decided not to: failure would only dent its reputation, muses Timothy Harris, who develops applied-science tools at the Howard Hughes Medical Institute’s Janelia Farms campus in Ashburn, Virginia. The other entrant is a team led by George Church at the Wyss Institute at Harvard in Boston, Massachusetts.
Grant Campany, senior director of the genomics X prize, hopes that other teams will step up to compete before the contest gets under way in September. Even if they don’t, scientists predict that its goals will be achieved within the next few years, whether through the prize or not. “If you could deliver that kind of performance you would have the commercial advantage by a large margin over anyone else,” Harris says. “That commercial advantage is worth way more than the X prize.”
Nature
 
497,
 
546–547
 
()
 
doi:10.1038/497546a

5:11 AM
29 May 2013
The sparse list of contenders for the Archon Genomics X Prize shows how far sequencing technology has come — and how far it still has to go. Barring any late surprises, only two teams will have signed up by the registration deadline of 31 May to compete for the US$10-million prize for the first to sequence 100 centenarians’ genomes in 30 days or less at a cost of $1,000 per genome (see page 546). The sequences must have no more than one in a million errors, be 98% complete and have correct haplotype phasing — a determination of which parent contributed each portion of a chromosome. It is not possible at present to meet this combination of goals with any single technology, but that does not explain why so few are reaching for it.
One reason why more teams are not lining up for the prize is that the promise of a genomic medical revolution is not being stalled by any lack of data. At genetic-medicine conferences,such as the University of California’s OME 2013 precision-medicine conference, held on 2 and 3 May in San Francisco, or the Big Data in BioMedicine conference held from 22 to 24 May at Stanford University in California, you will hear the same refrain. “We have more data right now than we know what to do with.” Figuring out how to interpret genetic data — and, more crucially, how to prove their value to patients and health-care systems — is the most pressing challenge in genomics today. Researchers can already sequence the protein-coding regions of a genome for less than $1,000. Getting more data on regions of the genome that they do not yet know how to interpret will not help to advance the goal of proving the medical worth of big data.
“The promise of a genomic medical revolution is not being stalled by any lack of data.”
Interpretation and analysis — making sense of the data — is now the real prize. Hence the launch of a spate of bioinformatics challenges as researchers compete to surmount that hurdle. They include Sequence Squeeze, a contest to develop the best sequence-data-compression algorithm; the Assemblathon, for the best program to assemble a genome sequence from scratch; the DREAM Challenges to analyse and predict biological interactions among gene products; the CLARITY genome-interpretation challenge; and contests at the annual Beyond the Genome meeting. Michael Schatz at the Cold Spring Harbor Laboratory in New York, who has curated many of these contests, is planning more challenges this year, including one at Cold Spring Harbor later this autumn. Bioinformatics contests have the advantage that they do not require physical manufacturing infrastructure, so they are more accessible to more would-be solvers around the world.

And it is very unlikely that anyone other than a well-financed lab or large company could attempt the current challenge. That also sets it apart from other competitions — the Google Lunar X Prize, for example — in which teams of professionals or even hobbyists can make a respectable showing. The thriving do-it-yourself biology movement, by contrast, cannot mount a credible challenge to the large life-sciences companies. The attempt is even beyond most biotech start-ups. The UK-based biotechnology company Oxford Nanopore, for example, which is trying to commercialize a promising technology pioneered by highly respected researchers, has raised at least $150 million in grants and investment since 2008 — but has yet to show that its technology can be used to sequence a complete human genome.There are other reasons why the genome X prize is a harder sell than other X prizes. The sequencing field is much more mature than were other industries that have been the focus of successful X prizes. Whereas there was no space-tourism industry before dozens of teams competed for the Ansari X Prize in 2004, for instance, there is already a thriving commercial market for sequencing. So any company that could meet the goals laid out in the prize already has its incentive — and it would be worth a lot more than $10 million. The value of the market leader in sequencing, Illumina of San Diego, California, is currently $8.8 billion.

That is not to say that the genomics X prize does not matter. The X Prize Foundation should be commended for revising the challenge, initially laid out in 2006, as the field evolved. It has also done a valuable service by working for two years with many partners, including Nature Genetics, to outline a judging scheme that can independently assess the quality and accuracy of a genome sequence and that is agnostic about the sequencing technology used. The foundation deserves kudos for prompting the field to reach farther; if past history is any guide, genomicists will reach that goal sooner than now seems possible.
Subscribe to: Posts (Atom)

MARI themes

{facebook#YOUR_SOCIAL_PROFILE_URL} {twitter#YOUR_SOCIAL_PROFILE_URL} {google#YOUR_SOCIAL_PROFILE_URL} {pinterest#YOUR_SOCIAL_PROFILE_URL} {youtube#YOUR_SOCIAL_PROFILE_URL} {instagram#YOUR_SOCIAL_PROFILE_URL}
Powered by Blogger.