HPCwire: Data Mining Algorithm Explains Complex Temporal Interactions Among Genes

Inphi Learn about LRDIMM http://lrdimmblog.inphi.com/

Since 1986 - Covering the Fastest Computers
in the World and the People Who Run Them

Translation Disclaimer

Subscribe | Sign In

Visit additional Tabor Communication Publications

Leading HPC
Solution Providers

June 22, 2010

Data Mining Algorithm Explains Complex Temporal Interactions Among Genes

June 22 -- Researchers at Virginia Tech, New York University (NYU), and the University of Milan, Italy, have created a data mining algorithm they call GOALIE that can automatically reveal how biological processes are coordinated in time.

Biological processes such as cell division, metabolism, and development must be carefully synchronized for proper cell function. How such events are coordinated in time is a complex problem in the field of systems biology. While researchers can gather temporal data about the activity of thousands of genes simultaneously, interpreting these datasets in order to understand higher order phenomena such as cell division requires the development of new analysis tools. The mathematically rigorous data mining algorithm GOALIE (Gene Ontology based Algorithmic Logic and Invariant Extractor) reconstructs temporal models of cellular processes from gene expression data. The researchers describe this algorithm in the early on-line Proceedings of the National Academy of Sciences (PNAS) to be released the week of June 21, 2010.*

The researchers developed and applied their algorithm to time-course gene expression datasets from the well-studied organism Saccharomyces cerevisiae, a budding yeast that is also used for raising bread dough and the manufacture of beer, wine, and distilled spirits. They applied their novel temporal logic-based algorithm to a range of yeast data sets involving cell division, metabolism, and various stresses. "A key goal of GOALIE is to be able to computationally integrate data from distinct stress experiments even when the experiments had been conducted independently," said Naren Ramakrishnan, professor of computer science at Virginia Tech, and lead author.

"GOALIE is part of a broader effort to combine data mining with modeling tools," said Bud Mishra, professor of computer science and mathematics with the Courant Institute of Mathematical Sciences at NYU, and corresponding author. Mishra, also a professor of cell biology with the NYU School of Medicine, is investigator on a $10 million National Science Foundation (NSF) Expeditions grant to develop novel computational reasoning tools for complex systems, focusing on biological organs to complex diseases as well as engineered systems. "GOALIE cannot just mine patterns but also extract entire formal models that can then be used for posing biological questions and reasoning about hypotheses," said Mishra.

A hypothesis in the yeast example is how genes organize into groups to perform a specific concerted behavior. "However, these gene groupings are not permanent, but shift as the cell begins orchestrating its next step. These transitions correspond to significant 'regrouping' of genes, which is indicative of a change in cellular state," said Richard Helm, associate professor of biochemistry at Virginia Tech, and co-author. Tracking down these transitions in time-based experiments is difficult, especially with thousands of genes changing in levels simultaneously. "When confronted with datasets this large we tend to focus on our 'favorite' genes or processes, leading potentially to a biased viewpoint," said Helm.

"GOALIE blends techniques from mathematical optimization, computer science data mining, and computational biology," said Layne Watson, professor of computer science and mathematics at Virginia Tech, and co-author. "It automatically mines the data in an unsupervised manner, identifying temporal relationships between groups of genes in order to gain a more unbiased and holistic understanding of time-based cellular behavior."

Specific strains of S. cerevisiae have been shown to have two robust biological cycles occurring simultaneously, namely the metabolic and cell division cycles. While the yeast cell division cycle has been well studied, its relationship to and coordination with metabolism are only now being worked out. GOALIE was able to recover the underlying temporal metabolic and cell cycle relationships in the datasets studied. "Through our temporal models, we have shown that S. cerevisiae reacts in a somewhat unified fashion, with cellular fate depending on core metabolism and cell division," the authors write in their paper.

"The metaphor that emerges from this analysis is that the metabolic state of the cell is essentially a fuel gauge, and there must be enough 'fuel in the tank' before permitting another key biological process, such as reproduction, to commence," said Helm. "The availability of energy controls whether a yeast cell divides or not."

"Our tools bring out the nature of temporal 'hardwiring' manifest in biological processes," said Ramakrishnan.

Helm adds: "In particular, they open up questions related to whether it would be possible to manipulate the system to adopt an aberrant cell state or make it proceed along a desired temporal order. The identification of well-defined states, such as found in hydrogen peroxide treatments, suggests that at this stage it may be possible to force the organism to adopt aberrant states."

For instance, the biotechnology industry currently employs microbes for a number of important commodity and specialty compounds, ranging from biofuels to pharmaceutical products. If cell division could be unlinked from metabolism, the microbial system would only need nutrients for maintaining metabolism, with fewer resources diverted to cell division. "This scenario would reduce overall bioproduction costs for the chemical of interest," said Helm.

Ramakrishnan, Mishra, and co-author Marco Antoniotti, associate professor of computer science with the University of Milan, are also inventors on a US patent application about GOALIE for which a notice of allowance has been issued. "We hope in the future our work can become key to understanding other important phenomena, like disease progression, aging, host-pathogen interactions, stress responses, and cell-to-cell communication," said Mishra.

*"Reverse engineering dynamic temporal models of biological processes and their relationships," by Ramakrishnan; Satish Tadepalli, a 2009 Ph.D. graduate of computer science at Virginia Tech; Watson; Helm; Antoniotti; and Mishra.

-----

Source: Virginia Tech

Share Options

Subscribe

Subscribe to HPCwire

Discussion

There are 0 discussion items posted.

Join the Discussion

Become a Registered User Today!

Registered Users Log in join the Discussion

RSS Feeds

Feeds by Topic

Feeds by Industry

Feeds by Content Type

Subscribe to All Content

All

February 10, 2012

Wolfram|Alpha Goes Pro

February 09, 2012

February 08, 2012

February 07, 2012

February 06, 2012

February 03, 2012

February 02, 2012

February 01, 2012

January 31, 2012

Feature Articles

AMD Opens Up Heterogeneous Computing

Last week, AMD used its Financial Analyst Day to talk up heterogeneous computing, the technology that the company is betting on to be the next "big thing" in the microprocessor business. To that end, company execs explained how their newly hatched Heterogeneous System Architecture (HSA) will evolve over the next three years to drive their product roadmap forward.
Read more...

HPC Funding Models Need to Encompass More Than Just the Purchase Price

Beyond the question of how much funding should be invested in high performance computing resources, it is also important to strive for the optimum funding model: how funding is tied to the service and how it enables and drives user behavior. As it turns out, these models are wrapped up in an IT culture that is often at odds with the way HPC is used.
Read more...

AMD Unveils New Strategy for Server Silicon

AMD is plotting a relatively conservative roadmap for its Opteron CPUs over the next year or two, even as it preps its heterogenous computing technology for the big leap into the server arena. At the company's 2012 Financial Analyst Day last week, AMD execs re-pledged their commitment to the server market and outlined a strategy that puts less emphasis on high performance cores and design complexity and more on power efficiency and building SoC products tailored to specific datacenter workloads.
Read more...

Read more HPCwire features...

Around the Web

Cray Yarcs Up Big Data

Feb 09, 2012 | Supercomputer maker goes after big data segment with new YarcData division.
Read more...

SGI Posts Q2 Loss as Revenues Rise

Feb 08, 2012 | Server and storage maker SGI delivered some unwelcome news to its investors on Wednesday.
Read more...

Microsoft Cranks its AMP

Feb 07, 2012 | Software maker offers heterogeneous computing in a C++ wrapper.
Read more...

The Power to Flop

Feb 06, 2012 | The race to exaflops could be one that the US is destined to lose.
Read more...

Cray Adds Baby Blue Waters to HPC Lineup

Feb 02, 2012 | New mid-range XE6/XK6 supercomputer configurations replace Xeon-powered CX machines.
Read more...

Read more headlines...

Introducing LRDIMM – A New Class of Memory Modules

01/17/2012 | Inphi | This paper introduces the LRDIMM, a new type of memory module for high capacity servers and high-performance computing platforms. LRDIMM is an abbreviation for Load Reduced Dual Inline Memory Module, the newest type of DIMM supporting DDR3 SDRAM main memory. The LRDIMM is fully pin compatible with existing JEDEC-standard DDR3 DIMM sockets, and supports higher system memory capacities when enabled in the system BIOS.

View the White Paper Library

Appro Chief Technology and Software Engineer Discusses the Appro Xtreme-X™ Supercomputer

The Appro next generation Xtreme-X™ Supercomputer recently launched has made headlines across the nation over the past month. In this interview we sit down with Appro CTO, Giri Chukkapalli to discuss the new system design of the Xtreme-X™ supercomputer to support future technologies. In addition, he also talks about the Appro Cluster Engine™ (ACE) management software suite, part of Appro’s cluster software stack that is tightly integrated with the new Xtreme-X™ Supercomputer.

Appro Video Blog: Interview with Sandia about Intel MIC Experimental Cluster

Doug Eadline, Editor of Cluster Monkey had a chance to sit down with Jim Ang, Technical Manager at Sandia National Laboratories for an interview about the “First of a Kind” Experimental Cluster, Appro Xtreme-X™ Supercomputer ,using Intel’s Knights Ferry (KNF) Software Development Platform for the Intel® Many Integrated Core (MIC) architecture. Just for the record, Knights Ferry is available only to select individuals including Jim Ang’s group at Sandia and represents a potential new direction in HPC.