Qlucore Bioinformatics Software: Using Proteomics to Understand Cardiovascular Disease

The Center for Interdisciplinary Cardiovascular - created in 2009 by the Brigham and Women's Hospital in Boston and the Japanese pharmaceutical firm Kowa Company Ltd - is looking into new ways to prevent and treat cardiovascular disease with a particular focus on proteomics. Since December 2014, it has been using Qlucore's Omics Explorer software to help analyze and visualize some of its proteomics experimental data.

Proteomics is the large-scale study of proteins (the proteome), contained within an organism, tissue, cell or organelle. Characterization of the proteome depends on mass spectrometry. As proteins are the primary effectors of cellular function, examining changes enables us to gain new insights into the cellular mechanisms and markers of disease.

CICS aims to use such novel discoveries in the development of first-in-class drugs that can modulate or regulate new protein targets responsible for the mechanisms for cardiovascular disease such as inflammation. The researchers also explore targets that lead to new modifiers of cardiovascular risks including high blood LDL cholesterol levels.

An ideal platform for data analysis

The proteomics group at CICS, led by Dr. Sasha Singh, integrates the latest technologies in mass spectrometry and collaborates with the center's main research groups. These groups' studies focus on four major biological processes contributing to cardiovascular disease: macrophage biology, lipid metabolism, arterial and valvular calcification, and metabolic disorders such as type 2 diabetes.

"When any of the three groups wants to look at protein-related mechanisms involved in their particular disease models, we can use all major global proteomics quantitative strategies to analyze the changes in abundance of anywhere between 1000 to 5000 proteins across a particular experimental time course or across various conditions," Dr. Singh explains.

"For instance, we have characterized the differences in abundance of the proteome across different tissue layers of a given organ. We have also monitored the changes in the proteome of cultured cells across a stimulation period. As a consequence, the number of samples to be analyzed can vary between three to over ten for a given experiment."

The researchers at CICS typically use cell lines with minimum heterogeneity such as the THP-1 human monocytoid cell line and RAW264.7 mouse macrophage-like cell line. Stimulations such as inflammatory agonists or antagonists are added and then the changes in the proteome over that stimulation period are measured. The researchers also use donor cells by, for example, isolating monocytes from peripheral blood of mice and human donors. Donor cells provide the realistic view of the monocyte proteome; however, donor-to-donor (genetic, epigenetic, or environmental) heterogeneity, most notably from humans, can translate to differences in responses by the proteome, making these studies more challenging to extract target proteins.

As a result, CICS' researchers work with the proteomics group to design experiments that minimize sources of variation due to experimental noise, thereby enhancing the biological variation they wish to pursue. Qlucore then provides an ideal platform for data analysis and visualization of these changes in the proteome that can be flagged as trends of interest for follow-up studies.

Currently Dr. Singh is working on over ten different proteomics projects, which involve the majority of the active research scientists at the centre, including those from Brigham and Women's Hospital and Kowa. "Most of the projects are creating big data sets of between 1000 to 5000 quantified proteins," she says.

Fast, intuitive and visually striking

CICS employs highly skilled informaticians who help develop novel in-house software to meet the unique demands of the center's research. However developing software is time consuming and often leaves little time for the informaticians to assist the biologists on simpler data analyses. Moreover, biologists are not often well versed in software codes and find working from the informaticians' toolbox very challenging. It eventually became evident to try Qlucore's software as an alternative means for data analysis at CICS.

"Qlucore is fast, very intuitive and the graphical user interface is simple to use. It means that the biologists at CICS can now do fundamental statistical analyses independently, having more control of their own data," says Dr. Singh. "It also means that our informaticians can concentrate on very advanced bioinformatics on the wilder frontiers of statistical analyses."

What has been interesting since using Qlucore, says Dr. Singh, is how the software has made it possible for the biologists to visualize their research results in striking ways.

"We want to understand the role and function of specific proteins of the lipoproteins' (LDL and HDL) proteomes. For instance, the HDL proteome has been described before and is known to comprise not more than 100 proteins. We, however, have investigated the proteome across various HDL fractions and demonstrated, with the visual aid of Qlucore-generated hierarchical clustering or heat maps, that the proteome is split into sub-proteomes; each sub-proteome identifies with a specific subset of HDL fractions," she explains.

CICS scientists have presented these data at several academic conferences and the HDL sub-proteome cluster output, generated by Qlucore, was very visually striking to the audiences, according to Dr. Singh. "The message was conveyed rather convincingly, and that is what we wanted," she says.

LDL cholesterol-lowering therapies using statins are a well-known way to reduce the risk of cardiovascular disease. Lower levels of HDL correlate with a higher chance of developing cardiovascular disease, but clinical trials on raising HDL have reported mixed results. What CICS is exploring is the hypothesis that the useful part of HDL is just a subset of the whole

A permanent fixture

Approximately half of the research projects at CICS have generated large proteomics datasets (greater than 1000 proteins) and have incorporated Qlucore into the data analyses steps. Since the first peer review process occurs in the laboratory itself, the researchers heavily depend on elements such as principle component analysis and hierarchical clustering to demonstrate to their peers in a group-meeting environment, the significance of their findings.

By doing so, the Qlucore has maintained itself as a permanent fixture at CICS for data analysis and representation, and communication of ideas and findings. The graphical and statistical outputs are also poised for publication-ready figures.