Discover the hidden patterns of the immune system with artificial intelligence
The immune system stores huge amounts of information about health and disease. Researchers are trying to decipher this information using artificial intelligence, with the aim of developing new diagnostic and therapeutic methods.
There is still a lot we don’t know about how the immune system works. We divide the immune system into different parts. One of them is the adaptive immune system. It develops during our lifetime. This part of our immune system stores information about all illnesses and infections a person has or has had. This information is stored as complex patterns in pathogen recognition structures, also called immune receptors. These are located on the surface of adaptive immune cells.
We can think of these patterns as the memory of the immune system. Models are instruction manuals that tell the immune system how to attack various infections and diseases. However, no one knows what these models look like. The potential for what they can tell us about the inner workings of the immune system, as well as the development of diagnostics and therapies, is enormous. So how can we find the patterns?
This is where machine learning comes in. Machine learning is a form of artificial intelligence. With the help of machine learning, we can let a computer discover previously hidden patterns for us.
Associate Professor Victor Greiff from the Institute of Clinical Medicine and Professor Geir Kjetil Sandve from the Institute of Informatics work there. Together with PhD students Milena Pavlović and Lonneke Scheffer, they developed the ImmuneML software for this purpose.
– We can use machine learning to find disease-relevant immune patterns, without knowing what they look like or what characterizes them. That’s what’s so unique and exciting about machine learning, says Greiff.
Immune patterns can tell us if a person is healthy or sick
The patterns that the immune system has stored can tell us about a person’s past and/or current illnesses or infections. Greiff, Sandve and their colleagues are now trying to determine which models belong to which diseases and infections. If they manage to find it, it may provide new and important insights into adaptive immunology.
Above all, it can also facilitate the diagnosis of various diseases as well as the development of new therapies.
“If we can find the patterns, we might be able to diagnose a number of diseases with a single blood sample,” Greiff says, adding:
– Patterns can tell us if the person is healthy or sick, and what disease(s) the patient may have.
The goal is to find the patterns of thousands of diseases
Greiff and Sandve collaborate with researchers and clinicians from various medical disciplines. They are currently trying to find the patterns of celiac disease and type I diabetes. For this, they are collaborating with researchers from UiO and the University of Florida.
– The real value will unfold when we learn the pattern of many diseases. In principle, you can then diagnose thousands of diseases from a single blood sample. That’s the goal, Sandve points out.
Can we use machine learning to find the pattern of COVID-19?
If you want to find something, but don’t know what you’re looking for, you have a difficult task ahead of you. Let’s say we want to find the pattern for COVID-19. How do we know which pattern it is among the millions of patterns that the immune system has stored?
It’s like looking for a particular snowflake among millions of other snowflakes. Without knowing what the snowflake we are looking for looks like.
However, with machine learning, that becomes a whole different matter. Then we can first let the computer find the patterns of a person with confirmed COVID-19 infection. This way the computer “learns” what the pattern of COVID-19 looks like.
– This is how machine learning works. First we have to teach the computer what is what. We can do this by letting it find the patterns of a person we know is healthy and a person we know has a specific disease, Greiff says.
ImmuneML will make machine learning accessible to more researchers
Initially, Sandve and Greiff’s plan was to develop software that would facilitate the application of machine learning in their own research. However, they soon realized that there was a need for common software in the field.
“It was a bit like having to dig out the site of the new building by hand each time, before we could start the construction itself, that is to say the analyses,” says Sandve.
Greiff explains that previous studies that applied machine learning were exploratory in nature. They were also poorly standardized and therefore difficult to reproduce.
– PhD student Milena Pavlović was tasked with developing software foundations that would allow her future studies to be conducted in a more efficient and reproducible manner than was typical in the field. When Sandve and Greiff saw how well the platform came together, they thought, why not share it with others? A second doctoral student, Lonneke Scheffer, also joined the team and today, more than a year later, the platform is now available for the entire field.
– ImmuneML allows you to work more uniformly. We can now compare different studies and assess the value of different approaches.
The goal is that researchers with different levels of experience and expertise in bioinformatics and machine learning can use ImmuneML. The software consists of three different variants with different levels of complexity. It also includes a user manual explaining in detail how to use the software.
– We also have our own YouTube channel with explanatory videos. We hope that many researchers interested in this will start using ImmuneML, concludes Sandve.
Associate Professor Victor Greiff and Professor Geir Kjetil Sandve have been collaborating and working on the development of ImmuneML since 2018. Researchers in their labs work interdisciplinary and now collaborate on most of their research projects. USIT at UiO and Elixir Norway contributed to the development of the ImmuneML software.