Decoding Blood: An Open Search for New Ways to Treat Common Diseases
A blood sample can reveal a lot about an individual’s health. For instance, reduced insulin levels can indicate diabetes, while other protein levels can point to heart disease. At the Genomics Institute of the Novartis Research Foundation (GNF), we think blood has a bigger story to tell about the causes of diseases and how they might be treated. “The great thing about blood is that it’s a sample of all tissues,” says John Lamb, Director of Bioinformatics at GNF. “If there was one place you’d look for global regulation of disease in the body, blood is an obvious choice.”
The approach we’re taking is called integrative genomics. It involves analyzing the blood serum of thousands of patients using a protein panel that identifies thousands of different proteins. We are correlating the variations in serum protein levels and variations in genomic sequences with the signs and symptoms of disease. “Our goal is to find new drug targets in data-driven and open-ended ways,” says Lamb.
Through this analysis we’re finding that patterns of protein levels in serum are directly related to disease states. Certain proteins can be found at atypically higher or lower levels in individuals with a disease compared to those levels in healthy individuals. “If a protein of interest also causes disease, we may be able to treat that disease by targeting the protein with drugs that lower or raise its levels,” says Will Barnes, a senior investigator on the project team.
Our goal is to find new drug targets in data-driven and open-ended ways.
We are focusing initially on diseases that tend to progress with age, such as obesity, diabetes, cardiovascular disease, and muscle wasting. The blood samples we’re working with come from a very large dataset along with extensive clinical data describing the medical condition of these individuals. To further expand the dataset, we also have arranged with several academic centers to work with blood samples and clinical information from individuals with chronic obstructive pulmonary disease, idiopathic pulmonary fibrosis, and heart failure.
Protein levels in blood are read out using a unique sensor panel technology. The panel initially detected approximately 1,000 proteins, but has been expanded fourfold through a collaborative effort led by GNF scientists. “This is the first time we’ve had both the power of modern genomic sequencing and the technology to look at so many proteins simultaneously in a single sample of blood,” says Barnes. “It’s a timely initiative.”
We have found that there are many proteins that are atypically elevated or lowered in people with a particular disease. “No one has ever looked at this number of proteins in the blood serum of people before,” says Lamb. “We were not sure what to expect, honestly, but what we’ve found is amazing. The relationship between proteins and disease is very structured.”
Our next steps involve validating the most promising proteins to determine if they are indeed drivers of disease and not just byproducts of a disease process. To do this validation, Tony Orth, Associate Director of Genomics at GNF, is leading an effort to generate the candidate proteins so that Barnes and his team can investigate their biological effects in animal and cell models. “If we find that a protein of interest drives a disease of interest, then we can start a drug discovery program,” says Barnes.