Entries by Thomas Pedersen

A few weeks ago, my wife and my youngest son heard what appeared to be a lunatic shrieking in our backyard. They rushed out the door to find me standing in the midst of about a dozen freshly dug holes, waving my arms and yelling at our mud-covered, 10 month-old Golden Retriever.

The dog, who had obviously excavated all of the holes, was simultaneously wagging her tail and sneezing the dirt out of her nose, utterly unimpressed with my tirade. My wife and son managed to get me inside with the promise of a cold drink and a warm compress, all the while calmly listening to me carry on about ruined landscaping and holes all the way to China.

Once I had calmed down and my wife had taken the dog off to the tub to wash away the evidence, my son solemnly informed me that it would be impossible for Gracie to dig a hole to China, because there were lots of rocks in the way and, besides, the center of the earth was way too hot. He also reminded me how much fun it can be to dig.

He was right.

It can be a lot of fun to dig. Now, I’m not talking about the back-breaking, working-on-a-chain-gang type of digging, but rather the treasure-hunting or the let’s-see-what-we-can-build type of digging. Just look at a bunch of kids at the beach or in a sandbox, and you’ll see what I mean. Old blues singers ask “Can you dig, it?” Peter, Paul and Mary tell us that they “dig” rock & roll music. Paul McCartney planned to “dig the weeds” when he turned 64 (somehow, I don’t think that ever happened.). The seven dwarves sang about “dig, dig, digging” in the mines where a million diamonds shine. The pirates of yore and modern-day archaeologists spend their days digging for buried artifacts and other treasures.

Digging can be pretty cool!

As genomics scientists, we also spend our days digging. Faced with a mountain of data, we dig and we sift and we dig some more, all the while looking for the clues buried in those mountains. Without the right tools, the task can be daunting; in fact, like treasure hunters without a map, we could dig forever without ever finding anything of value.

Here at SAS, we make the tools that make the digging easier. For example, new features in JMP Genomics 4.1, which is due out later this year, include interactive, graphical tools that will allow you to visually evaluate and compare complex statistical results for thousands of genes and other markers across an entire genome. Once you’ve identified one or more genes of interest, other new features make it easy for you to annotate those genes and link your analysis to online data bases. These tools won’t help you dig to China, but they will help you sift through your data until you find the treasure.

Excuse me, now. I have some digging to do.

Now that JMP Genomics v4.1 has gone to production, I have found a little time to catch up on my reading. As I was perusing a recent issue of the Proceedings of the National Academy of Sciences (USA), I was struck by the fact that the traditional limit of 5 pages has been scrapped and that articles of 6 or more pages have become common. Papers have grown as investigators try to pack as much information as they can into more and more space. Even authors of Science reports, which have traditionally been limited to 2500 words, get around page limitations by regularly including links to Supporting Online Materials, web sites that contain vast amounts of additional data and descriptions. All of this information tends to overwhelm the reader.

Perhaps I’m just showing my age again, but I miss the days when you were limited to a strictly-enforced set number of words or pages and there were no Supplementary Online Materials. Authors had to ensure that all of the supporting facts and figures needed to tell a good story were included in the article. These requirements forced authors to carefully consider which pieces of data were the most important and to use the most concise language possible. Good papers were razor sharp and told you just what you needed to know. They were a pleasure to read.

“Just the facts, ma’am…”

On the old Dragnet TV series (I’m most definitely showing my age here!), Joe Friday used these words to cut through all the fluff and get down to the essentials. We follow the same philosophy when we write the documentation for JMP Genomics. We have divided the JMP Genomics User Guide, by theme, into nine different volumes. We describe each process in its own chapter. We tell you what the process does and what you need to run it. We then show an illustrative example and tell you how to interpret the results.

Each chapter is structured in the same way; once you learn how it works for one process, you know how it works for every process.

A colleague of mine, who documents a different software package, recently told me that the JMP Genomics documentation was rather simple. Rather than being insulted, I took her comment as the highest form of praise. Software documentation should not contain a lot of fluff. Instead, it should just tell you what you need to know in an easily accessible format and then it should get out of your way.

The JMP Genomics User Guide is designed to do just that.

Sometimes, I’m totally astounded at how much our science has advanced since my days as a graduate student. Back then, the closest anyone got to “genomic” studies of eukaryotic organisms involved “melting” DNA and watching it come back together using CoT curves. Cloning and sequencing a single cDNA could get you a paper in Science or even (if it was especially important) Nature. As a classically trained molecular biologist, I was used to thinking about single genes. We ran northern blots and probed them with single cDNA probes. On a good day, our sequencing gels could resolve up to 300 nucleotides, maybe a few more if you did multiple loadings. Experiments were labor-intensive and not very quantitative. While we could generate some very pretty pictures, we certainly couldn’t do statistics. Back then, our science was limited by a lack of data.

Since that remote time (half-way back to the Pleistocene, as my kids would say), we have made incredible progress. We have sequenced the genomes of a growing list of diverse organisms. We can quantitatively assess the expression of not just one gene, but of every gene in an organism, all at the same time. We can ask global questions that we could never have asked just a few short years ago, and we can get answers to those questions in a relatively short period of time. In fact, today we have the opposite problem: far too much data! We used to spend months or even years gathering a few crucial data points that could be assessed by a mere glance at an autoradiogram. Today, we can do an experiment in a fraction of the time, but the analysis takes so much longer. Fortunately, our tools and skills are evolving along with our science.

The initial release of JMP Genomics in 2006 married the visualization capabilities and ease of use of JMP software with the power of SAS. It offered researchers more than 100 different processes for importing, manipulating and analyzing the vast amounts of data generated by the new technologies. The recent release of JMP Genomics 4.0, builds on an already strong platform of data management and analysis tools. We have added features to and enhanced the power of all of the existing processes. In addition, we have added 16 totally new processes. In fact, this latest release contains almost 200 different processes for importing, assessing, normalizing, annotating, and exploring genetic and microarray data. Every process is fully documented and available for you to use as is or to adapt to your particular needs. You can modify existing processes and workflows or build new ones and add them to your menus. In addition, if there is something special that you need, just let us know, and we’ll work with you to build it. As always, we remain committed to helping you meet your research goals.

We have come so far in such a short time. Where will our science go next? You will help decide, and JMP Genomics will help you take us there. We’re already hard at work on our next release. Stay tuned. The best is yet to come.

Our goal at JMP Genomics has always been to make it easier for you, the genomics scientist, to analyze and interpret your data. The JMP Genomics User Guide has been there to help you choose which analytical process(es) best meet your needs by showing you what you need to run each process and, through the use of an illustrative example, how to evaluate and interpret your results. This documentation has grown and evolved as JMP Genomics has grown and evolved. We’re proud to announce and make available to our users the newest edition of the JMP Genomics User Guide.

This update to the JMP Genomics User Guide combines and extends the best features of the original JMP Genomics User Guide and the JMP Genomics User Guide – Supplement that you received with JMP Genomics. Existing chapters have been revised, and new chapters have been added. All chapters have been fully updated with detailed illustrations and explanations for our current version, JMP Genomics 3.2.

The updated JMP Genomics User Guide is divided among the nine different volumes listed below:

• Getting Started with JMP Genomics. This manual lists installation instructions and requirements, provides a brief introduction to JMP Genomics and this documentation, and detailed descriptions and instructions for setting up and using the basic JMP Genomics workflows. This manual also provides a troubleshooting guide, various appendices, a comprehensive glossary and a complete reference list for all of the manuals.
  
  
• Designing Experiments with JMP Genomics. This manual describes how to use JMP’s DOE functions to design your experiment and generate an experimental design file, which is the first step in your analysis.
  
  
• Data Import and Manipulation with JMP Genomics. This manual provides detailed descriptions of the different types of data sets used by JMP Genomics, how to import data from different sources, and how the resulting data sets can be manipulated and transformed into formats appropriate for specific analyses.
  
  
• Genetic and Copy Number Analysis. This manual provides detailed descriptions for manipulating genetic data sets, determining specific marker statistics, carrying out various association tests, assessing linkage, mapping quantitative trait loci, and performing haplotype and copy number analyses.
  
  
• Microarray Analysis. This manual lists and describes a multitude of procedures for performing quality control on your data, normalizing it, looking for patterns, fitting statistical models to the observations and comparing the aggregated results.
  
  
• Predictive Modeling. Can your data be used as the basis for making predictions of future experiments? This manual details a number of different predictive models as well as procedures for comparing the relative efficacy of the different models with your data.
  
  
• Annotation Analysis. Once you have identified one or more significant genes, you must determine the biological significance of those genes. This manual provides tells you how to use JMP Genomics tools to incorporate biological meaning with your statistical results.
  
  
• Spectral Preprocessing and Analysis. This manual describes the JMP Genomics processes that are useful for analyzing two dimensional and three-dimensional spectra.
  
  
• Programming Guide. This manual provides an introduction to building and programming your own analytical processes. Processes you create can be added to the JMP Genomics menu.

To access this updated resource, just download the zipped file from the JMP Genomics Web site, unzip it and copy it into your JMP Genomics documentation folder. After you install it, you can access all nine volumes as before: Just select Genomics > Documentation and Help > User Guide to go to the User Guide entry page. From this portal, just click on the appropriate link to go to a specific manual. Each manual is dynamically cross-referenced to each of the other manuals – just follow the links embedded in the chapters to access the additional information you need.

Registered users of JMP Genomics should watch their Inboxes for an e-mail containing instructions on how to download and install the latest update to the JMP Genomics User Guide.

Whew! I’m glad that’s over. It was really crazy for a while. Writing detailed documentation for over 30 new processes plus documenting all the updates to existing processes really kept me busy. But it was worth it! Not only did we add a lot more features to JMP Genomics (including a whole new section on copy number analysis) but we also made changes that make JMP Genomics more powerful and easier to use.

The documentation for JMP Genomics continues to evolve. Our goal is to make it as easy as possible for our users to choose a process, identify what they need to run that process, discover all of its features, see a practical example and evaluate their results. Currently, over 40 new chapters do just that. We will continue to add to that documentation as new processes are added to JMP Genomics and, at the same time, adapt existing documentation to this format.

JMP Genomics 3.1 is out the door and I can relax a bit. Wait a minute, didn’t the developers say they were starting to work on 3.2? Oh well, back to work…