A team of scientists at the Institute of Human Genetics of the Christian-Albrechts-University and the University Medical Center Schleswig-Holstein in Kiel, Germany, is currently using sophisticated data analysis software on a number of national and international projects in order to study the epigenetic alterations related to several cancers, including malignant lymphoma, colorectal cancer, and hepatocellular carcinoma, as well as developmental disorders and other diseases.
The team, led by Dr Ole Ammerpohl, is currently studying the raw data obtained from array-based DNA methylation analysis, including Illumina’s HumanMethylation27k Bead Chip, which provides semi-quantitative data from more than 27,000 CpG loci.
Although studies like these are proving invaluable to the study of human biology, the amount of data that is produced by this kind of research is enormous. As a result, it is impossible to derive any real biological meaning from these findings unless sophisticated analysis methods are used to help interpret the data effectively.
“Because of the amount of data being analysed, conducting microarray analysis has always been a hassle,” says Dr Ammerpohl. “Larger studies, especially those which include multiple samples that need to be analysed on comprehensive array platforms, have traditionally been very time-consuming, and have also required a considerable amount of computer power. Plus, in most cases, it is mandatory to define the exact conditions before beginning the intended analysis, and just this process alone can take hours. Worse still, any recalculations – even if they are just minor adjustments to these initial conditions – would mean that the whole process would have to be restarted from the beginning.”
Fortunately, new technological advances in this area are making it much easier for scientists to compare the vast quantity of data generated by epigenetic studies, to test different hypotheses, and to explore alternative scenarios within seconds. As a result, the latest generation of data analysis software is helping scientists to regain control of this analysis, and to realise the true potential of research in this area.
Epigenetics: unlocking the mysteries of human disease
Epigenetics, the name given to the study of changes in phenotype (appearance) or gene expression caused by mechanisms other than changes in the underlying DNA sequence, is helping to provide a link between the (static) genome and the unsettled environment, and therefore to establish a gene expression profile that will enable the survival of the cell in certain environmental conditions.
“The methylation of the cytosine residue in a CpG dinucleotide is probably the best characterised epigenetic modification in the human epigenome,” explains Dr Ammerpohl. “DNA methylation is essential for normal development; it is not only a key player in gene expression control and parental imprinting, but it also assures genomic integrity by preventing repetitive sequences from recombination. At the same time, it also silences parasitic sequences, such as retroviral sequences which have been integrated into the genome.”
Epigenetic variations are involved in many processes, starting with transgeneration heritable effects, but also in areas like stress tolerance, behaviour, psychological problems, drug addiction, and severe developmental disorders and diseases. Since epigenetic modifications are more responsive to drug treatment than genetic alterations, understanding epigenetic alterations might provide new therapeutic options in the future.
As such, Dr Ammerpohl, along with his team of four scientists and three technicians, is working to understand how alterations in the DNA methylation pattern contribute to a specific phenotype or disease like cancer, as well as the way in which tumour induction introduces epigenetic alterations which support tumour growth and progression.
Coping with information overload
Until now, most of the software that has been designed to study areas like array-based DNA methylation has mainly focused on the ability to handle increasingly vast amounts of data, which means that the role of the scientist/researcher has been largely set aside. As a result, a lot of data analysis has been passed on to bioinformaticians and biostatisticians.
A new generation of data analysis software is helping to redress the balance, however, and is already playing a key role in unveiling important new discoveries, since it allows the actual researchers involved to study the data and to look for patterns and structures, without having to be a statistics or computer expert.
At the same time, the overall performance of data analysis software has been optimised significantly over the past three years. With key actions and plots now displayed within a fraction of a second, researchers can increasingly perform the research they want and find the results they need instantly.
“For us, a big advantage of the latest data analysis software is its speed,” says Dr Ammerpohl. “We are using an application called Qlucore Omics Explorer, and this software makes it very easy to assign samples to defined groups, to change the applied statistical methods, to create new groups, and to modify the thresholds for items such as variance and p-value in real time, with results returned immediately.”
Data analysis software rises to the challenge
Products like Qlucore Omics Explorer are now making it possible for scientists to analyse proteomic, genomic and microarray data with a combination of statistical methods and visualisation techniques such as heat maps and principal component analysis (PCA).
As a result, scientists studying DNA methylation analysis and other genomic data can now analyse all of this important information in real-time, by themselves, directly on their computer screen, since the software can provide instant user feedback on all actions, as well as an intuitive user interface that can present all data in 3D.
By using Qlucore Omics Explorer in this way, Dr Ammerpohl is able to apply different statistical approaches, and to keep track of the effects in a PCA or cluster analysis. As such, subgroups in the sample collection – comprised of specific groups of genes – can be identified intuitively. All relevant statistics (together with the corresponding variable and sample list) can be exported, as well, so that they can be easily integrated into publications or presentations.
Making sense of important data
The latest data analysis software can generate PCA-plots between various sample data interactively and in real time, directly on the computer screen, and work with all annotations and other links in a fully integrated way, all at the same time. This approach has helped to open up new ways of working with data analysis and, as a consequence, has helped the biologists to be more actively involved in the analysis process.
Qlucore Omics Explorer, for example, graphically represents the high dimensional data in the form of 3-dimension plots on the computer screen. This instant visualisation technique is then combined with powerful statistical methods and filters, all of which are handled with just a single mouse-click.
“As humans, we are all used to interpreting 3D pictures in our environment, and so our brain is able to find structures in complex 3D figures very quickly. Therefore, it’s no wonder that a 3D presentation of complex mathematical/statistical coherences makes its interpretation much easier for us,” Dr Ammerpohl adds.
A step-by-step approach to data analysis
For their research into array-based DNA methylation, Dr Ammerpohl and his colleagues have compared the DNA methylation pattern in normal liver tissue, in cirrhosis of the liver, which is thought to be a precursor of hepatocellular carcinoma (HCC), and in HCC. In addition to the DNA methylation values obtained from the array analysis, the researchers have also included information on the tissue, clinical features or exogenous exposures like viral infection of the patient.
Afterwards, the analysis of the data could easily be performed. By selecting the appropriate test and adapting the applied statistical thresholds for p-value, false discovery rate, or the minimal variance by using sliders in the Qlucore software, groups of samples with similar characteristics could be identified easily by PCA.
Furthermore, a PCA of the variables (in this case, the DNA methylation values of the CpG loci) or a hierarchical cluster analysis is available. In this particular study, the team of researchers identified genes which acquire epigenetic alterations already in cirrhosis, which are maintained in HCC, or genes acquiring epigenetic alterations in HCC exclusively.
The future
Dr Ammerpohl has been using Qlucore Omics Explorer to analyse numerous epigenomes of tumour entities including colorectal cancer, HCC and lymphomas, as well as of developmental sex disorders or imprinting diseases. These studies have already resulted in valuable data that is helping Dr Ammerpohl and his team in their efforts to understand the epigenetic background of these diseases or disorders.
“The exceptional speed that Qlucore’s software provides is very important for us. The fast analysis of the data highly contributes to the identification of subpopulations in a sample collection or a list of variables,” says Dr Ammerpohl. “Without a doubt, these rapid results – and the way in which they are presented – prompted us to perform analyses that we would have never performed otherwise.”
“We don’t want to give away too much at the moment, as many of our major findings have yet to be published, but I can report that we are feeling very positive about our research in this area, and about the future of the entire project,” he adds. “It is fair to say that we have already obtained some very interesting results.”