May 2014
Revealing human-growth patterns in transcriptomic data
Making sense of childhood growth
Background
Dr Adam Stevens, is a human development systems biologist working in Professor Peter Clayton's group at the UK's University of Manchester. In addition to industrially-funded and commercially-sensitive research into drug responses, Dr. Stevens and colleagues have published a number of papers on the pharmacogenomics of growth hormones.
Young humans grow in interesting ways. Dr. Stevens points out that humans are the only animals that return to rapid growth in puberty. Other animals - even other mammals that have a long juvenile phase - grow rapidly after birth and then their growth rate slows down until eventually tailing off as they reach their adult size.
Such observations go beyond quirky facts. Understanding how children grow is key to understanding many childhood diseases and conditions, particularly those affecting growth.
"An active return to growth is a biologically dangerous thing to do. If you want to understand the impact of drugs and disease on childhood you need to understand this," Dr. Stevens notes. "A child is very different at two than at four or 10 but medically they are treated the same."
His research aims to bring more understanding to the process. In addition to bench studies, he carries out in silico studies on a range of omics datasets derived primarily from blood samples throughout human childhood. "The problem is that there is no suitable animal model for this so to study the active return to long bone growth the only model we can use is humans."
Technical situation
Steven's datasets include transcriptomic data - which measures RNA and microarray RNA sequencing. He also looks at the integration of genetic data and omic data (such as proteomics and phosphor proteomic data).
He uses such datasets to study both normal childhood growth and growth with various impairments. For example, he looks at short stature, where there is a growth hormone deficiency, and chromosomal syndromes, for example Turner syndrome, where girls are often treated with growth hormones. He also looks at diseases in children where growth has an impact, such as cancers. "Many cancers in children have a specific onset age, which suggests that different sets of genes are involved," he explains. Some drugs, such as steroids, also affect growth.
"Now we understand better how to map growth from a molecular level and its impact on pharmaco genetics," explains Dr. Stevens. "The methods we've developed allow us to study it more clearly, with highly-resolved molecular maps."
Solution
To assist with this research, he turned to theQlucore Omics Explorersoftware from Sweden-based Qlucore. Dr. Stevens found that a fundamental problem when running a project was that you might need a statisticians but they don't always understand the biology. One of the key drivers in usingthe softwarewas the visually-driven interface. Data needed to be presented in a way that could be understood and analysed by reviewers who are not experts in statistics and mathematics.
"The sheer genius is letting the visualisation drive the analysis. We want to have software that's accessible to non-statisticians. A fundamental problem that I see all the time is if you're running a project and realise you need statistics expertise you find statisticians but they don't understand the biology. There is a communication problem."
Dr. Stevens imports different microarrays in a range of conditions into the tool and tests whether there are differences or not. The industry-standard transcriptomic datasets might include around 400 microarrays, each with 55,000 data points. He has also imported other omics data into Qlucore. These datasets, he says, tend to be smaller but still have thousands of data points per sample.
What he does with this data depends on the data and the question he wants to ask.
"If I come in with a hypothesis, it's simple to use and replicates what lots of other tools do," he notes. "The power of Qlucore is in assessing structure when I'm coming in with hypothesis-free data."
In fact, he observes that several times he has been able to spot previously undiscovered patterns also in published datasets because of the visualisation capabilities of the software.
END
About Qlucore
Qlucore started as a collaborative research project at Lund University, Sweden, supported by researchers at the Departments of Mathematics and Clinical Genetics, in order to address the vast amount of high-dimensional data generated with microarray gene expression analysis. As a result, it was recognised that an interactive scientific software tool was needed to conceptualise the ideas evolving from the research collaboration.
The basic concept behind the software is to provide a tool that can take full advantage of the most powerful pattern recogniser that exists - the human brain. The result is a core software engine that visualises the data in 3D and will aid the user in identifying hidden structures and patterns. Over the last few years, major efforts have been made to optimise the early ideas and to develop a core software engine that is extremely fast, allowing the user to interactively and in real time instantly explore and analyse high-dimensional data sets with the use of a normal PC.
Qlucore was founded in early 2007 and the first product released was the "Qlucore Gene Expression Explorer 1.0". The latest version of this software, now called "Qlucore Omics Explorer 2.0", was released in May 2009, and represents a major step forward with the added support for hierarchical clustering, scatter plots and powerful log function. The combination of instant visualisation and advanced statistics support gives the user new opportunities. All user action is at most two mouse clicks away. The Company's early customers are mainly from the Life-science and Biotech industries, but solutions for other industries are currently under development.
One of the early key methods used by Qlucore Gene Expression Explorer to visualise data is dynamic principal component analysis (PCA), an innovative way of combining PCA analysis with immediate user interaction. Dynamic PCA is PCA analysis combined with instant user response, a combination which provides an optimal way for users to visualise and analyse a large dataset by presenting a comprehensive view of the data set at the same time, since the user is given full freedom to explore all possible versions of the presented view. Later versions combine PCA analysis with other analysis methods such as hierarchical clustering.