Article Index

Calendar

CD's
- Cost Justification CD
- LIMS Primer CD

Email Newsletter

Featured Article

Guest Book

Industry News

Introduction to LIMS

LIMS Products & Services

LIMSource

Reprints

Travel & Learning: Is Your Presence Required?

2009 is going to be a difficult year for conferences, seminars, and short courses looking for attendees. In this time of restricted budgets, spending money to attend an off-site event has many people wondering if it's worth the expense and whether they can't get the information they need some other way, such as on the Internet.

Unfortunately, the Internet rarely offers the depth of detail necessary to extend one's knowledge in the way that a short course or seminar provides. And certainly there's nothing quite like the highly subjective encounters that an exhibit hall offers to look and touch and test drive various instruments and software as well as meet the people who will support those solutions. A picture of a Mass Spectrometer may be worth a thousand words, but actually looking at the instruments, being able to compare various makes and models at the same time, and talking to their representatives in a neutral environment is priceless.

The difficulty lies in the number of venues available and selecting the most worthwhile events that can be justified and budgeted.

See Something

The venerable Pittsburgh Conference (Pittcon) has been steadily attracting fewer and fewer attendees. The number of exhibitors remains high, but the quality of the symposiums is uneven. The very technical sessions still remain the purview of the scientist, but the more general sessions -- particularly the Informatics sessions sad to say -- are largely presented by vendors pitching their software and services. While this saves a trip to their booth, the information provided is sometimes misleading and must be scrutinized. It is still worth going to Pittcon to attend the more in-depth scientific sessions that feature the latest experiments and research and to visit the exhibit hall. Certainly no other conference in North America collects as many vendors in one place. The next Pittcon will be held March 6-15 in Chicago, IL. For details, visit http://www.Pittcon.org.

Learn Something

But the main reason to venture forth from the lab and office is to learn something, not go shopping. Thankfully, there are a number of very targeted small conferences that can provide detailed information in very specific applications. For instance, Cambridge Healthtech Institute hosted two back-to-back conferences last September of great interest to researchers in genetic analysis. Next-Generation Sequencing Data Analysis and Exploring Next-Generation Sequencing were held at the Rhode Island Convention Center in Providence, RI from September 21-25. The sessions were typically half and hour long, which enabled the speakers to provide a reasonable amount of detail. Short courses were held before the conferences, a small exhibition hall highlighted about 20 vendors, posters were presented, and break-out luncheon table discussions of particular topics such as Accurate Normalization for Accurate Gene Expression and Bioinformatics Pipelines for Metagenomic Studies were provided to promote networking.

These particular conferences addressed the on-going issue of data management that all laboratories face. Now, managing the terabytes of data and turning new sequencing data into knowledge has given rise to new specialties such as biostatisticians. There are orders of magnitude more data to cope with while at the same time the focus is now on the individual human via whole genome mutational profiling. The first keynote speaker, Gabor Marth, Assistant Professor at the Department of Biology for Boston College, discussed the pros and cons of various technologies that address this challenge as well as the representational biases of each machine; i.e., the Roche 454, Illumina's Solexa, Applied Biosystems' SOLiD, and Helicos Biosciences' HeliScope. Marth drew upon his research experience in the academic environment to point out the amplification errors to expect from each instrument and how to accommodate those. He also detailed the error rate overall for each machine. For instance, he found that the Solexa overall error rate is about 1 percent and that most of the errors are substitutions with a very small fraction being insertions and deletions. The Roche 454 error rate, on the other hand was also very low - about .053 percent - but that the errors that do occur tend to be insertions and deletions. This information is important to know to ensure the accuracy of the data and the researcher's conclusions. Is the vendor going to tell you this? Probably not, or probably not in this way. You'll only find out accidentally, or by doing a detailed examination such as Marth and his co-workers did, or by attending a seminar where your industry peers share the knowledge.

Hear How to Do Something

Marth talked about much more than the instruments. He discussed applications from genome resequencing to transcriptions sequencing as well as industrial resequencing and what he called the "variation discovery toolbox" that enables researchers to perform base calling, map reading, paired-end read alignments, gapped alignments, SNP and short-INDEL discovery, structural variation discovery software and data visualization. March addressed the new analysis tools needed by industry. He covered data storage and data standards. He summarized by stating that "Informatics tools are already effective for basic applications but there is a need for further development. There is a need for genetic analysis tools and for specialized tools for specialized applications such as genetic profiling."

Subsequent sessions were equally informative. Jared Flatow an analyst/programmer from Northwestern University dealt with scalable bioinformatics, stating that "distributed computing is a great concept but it's hard and writing the debugging tools can be a chore." He pointed out that distributed programming languages are complex and proposed a simpler solution, a Map Reducer, that automates distribution and fault-tolerances. He offered other solutions that he has unearthed which help the researcher cope with data management and storage, including the interesting use of Amazon Web Services to create an elastic compute cloud (EC2) and simple storage service (S3) that utilizes another organization's servers.

In one way or another, all the speakers focused on solving the problem of transforming next-generation sequencing data into functional knowledge. Fully automated workflows were promoted. Alignment and annotations were highlighted. IT scalability and flexibility were discussed. The benefits and challenges of on-demand Informatics was examined. Whether all the information provided could be used was not as important as that a wide variety of solutions were presented.

The depth of the presentations was gratifying; the scope of the discussions ranged over all aspect of data management for this particular field. It's virtually impossible to achieve such an experience and absorb so much diverse information any other way. The next Next-Generation Sequencing Data Analysis conference will be held March 18-19 at the Hilton San Diego Resort in San Diego, CA. For details, visit http://www.healthtech.com.

Justifying Conference Attendance

The example about visiting an exhibit hall to look at different vendor's Mass Spectrometers highlighted above may not be the best example since those instruments have been on the market for some time. Next-Generation Sequencing instrumentation, however, particularly the newer solutions, still have very individual capabilities and quirks. Reviewing a number of such solutions, all in one or two days, while at the same time gaining insights into how to perform research using them more efficiently, is an extremely effective use of the researcher's time.

Hearing fresh ideas and innovative approaches to data challenges make the time and money spent attending specialized conferences worthwhile. Researchers buried under data and closely focused on the problems at hand benefit from interacting with their colleagues and gaining a fresh perspective on solving data management challenges.

Learning how to solve data management problems can lead to increased lab productivity, in turn leading to lowered laboratory operating costs and putting the researchers just that much closer to the next revenue-generating solution for their organization.