Last week I was able to attend two very interesting conferences about the latest developments in the analysis of sequence data. I started the week in San Diego at the Illumina iDEA Conference and finished it in Philadelphia at Children’s Hospital of Philadelphia’s (CHOP) Third Annual Next-Generation Sequencing Symposium. The Illumina conference was an opportunity to learn about the latest innovations in analysis tools for NGS data, while the CHOP symposium focused more on applied research and the exciting results that people are getting from sequence data right now.
The Illumina iDEA (Illumina Data Excellence Award) Conference was the culmination of the inaugural iDEA Challenge. The iDEA challenge was created by Illumina to encourage development of software tools for analysis of sequencing data. Interested parties were provided with RNA and DNA sequence data as well as methylation data for a set of eight subjects with the challenge to create novel algorithms and visualization tools to improve the overall scientific utility of the data. Prizes were awarded based on novelty and usefulness of algorithms and visualizations, as determined by an independent panel of judges (not directly affiliated with Illumina). Twelve finalists were selected to give presentations about their entry at the conference.
The entries covered a broad range of applications, but there definitely seemed to be an emphasis on tools for comparative genomics and structural variant analysis methods for small numbers of samples. One of the big challenges with visualization of sequence data is presenting structural variants in an intuitive way. Researchers are so accustomed to viewing genomic data in one dimension with data points aligned on a standard reference genome, that it is difficult to for us to process alternate arrangements, especially when comparing data across multiple samples and annotation sources. This particular question was addressed by several entries. We congratulate our colleagues at Enlis Genomics and Stephan Schuster’s “inGAP” team at Penn State for being judged as the best overall entries from the commercial and academic fields, respectively.
Aside from the presentations by the finalists, I was very interested in the opening address by Illumina CEO, Jay Flatley, and the keynote presentation given by Broad Institute Creative Director, Bang Wong. Flatley spoke about the current state of the industry from Illumina’s perspective and gave some background on the iDEA Challenge. He says that the major burden of sequencing work over the past decade or so has been generating the catalogs of reference data necessary to enable thorough analysis of the human genome. Now, that burden is shifting toward bioinformatics and analysis as more and more individual genomes are sequenced. This shift will require improved software tools, which was part of the motivation for hosting the iDEA challenge. Flatley also said that the cost of sequencing a whole genome at list prices is now below $5,000 and reminded the audience that the cost of array-based genotyping is falling at a similar rate. He said that Illumina’s individual genome sequencing service is growing quickly and may ultimately become their largest market. Consumers who order this service will soon start to receive their sequence data on an iPad (unfortunately, I wasn’t able to get my hands on the prototype).
Bang Wong gave a fascinating talk about the history of illustrations and data visualization in medicine and biology together with a discussion of the best practices for data representation. He demonstrated that the most effective visualizations are often the simplest and warned against overuse of color gradients and 3-D effects that can often be confusing or may even unintentionally mislead the audience. He pointed out that that other features such as size and motion can be very powerful but are underutilized. Wong has the unique ability to see things simultaneously through the eyes of a scientist and an artist, giving him a great perspective on data visualizations. (Yes, I took extensive notes on his presentation!)
After a cross-country flight, I attended the Third Annual Next Generation Sequencing Symposium at the Children’s Hospital of Philadelphia. This day-long event featured several speakers from CHOP and the University of Pennsylvania who spoke about their current research involving NGS technologies. After spending two days hearing presentations about methodology at the iDEA conference, it was both refreshing and exciting to hear about the results that people are getting right now from practical applications of similar methods. Topics ranged from identification of therapeutic targets for childhood cancers to identifying the cause of inherited eye diseases.
Richard McCombie from Cold Spring Harbor Laboratory gave an excellent keynote address about the challenges and opportunities presented by NGS as a disruptive technology. After listening to the day’s presentations, I was reminded of the lessons learned from last fall’s GAW conference – specifically, that there is no substitute for good annotation data when studying DNA sequences. This realization completed the circle for the week. The annotation data that makes sequence analysis possible has been meticulously curated throughout the previous decades as Jay Flatley told us in the opening address of the iDEA conference.
The future is very exciting, and I look forward to the many new discoveries to come as we continue to unravel the genome. …And that’s my 2 SNPs.
I was just wondering the other day – what you have been up to since graduation. So I was pleased to see this blog post! We’ve just recently purchased licensing for Golden Helix here at work. Excellent software! Keep up the great work! 🙂