Two non-profit organizations challenge cancer researchers to create prescriptive algorithms that can accurately suggest a personalized vaccine against tumor growth. It’s survival of the fittest algorithm. If the effort is successful it will be a critical strike against cancer, making this humanitarian effort one worth watching.
Technology is a tool to improve life. To do that, it tends to imitate life. For example, our devices resemble us by consuming fuel to operate, and many of those devices such as vehicles and industrial machines produce waste as well.
The tendency to integrate technology into natural environments is becoming more pronounced with the advent of Industry 4.0, so it is no surprise that researchers are looking to use the evolutionary principle of ‘survival of the fittest’ to attack one of the biggest medical problems of the modern world: cancer.
Two non-profit organizations, the Parker Institute for Cancer Immunotherapy in San Francisco and the Cancer Research Institute of New York City, are planning to pit prescriptive algorithms against one another to see which is best able to suggest a personalized vaccine from the DNA in a patient’s tumor.Two non-profit organizations, the Parker Institute for Cancer Immunotherapy in San Francisco and the Cancer Research Institute of New York City, are planning to pit prescriptive algorithms against one another to see which is best able to suggest a personalized vaccine from the DNA in a patient's tumor.Click To Tweet
Vaccinating a patient against cancer is a highly complicated thing to do. Any patient could have hundreds of cancer mutations, and if the correct mutation can be identified and made visible to T-cells (the ‘attackers’ of the immune system), it could provoke the patient’s immune system to eradicate cancerous cells, which would provide patients with a non-toxic way to fight cancer.
Framing the Algorithm Showdown
Melanoma and lung cancer are the best targets for this research because they have a large number of mutations. Additionally, Melanoma represents the vast majority of skin cancer deaths, and lung cancers are the single biggest killer among cancers, claiming one out of every four deaths. If scientists can find the key to immunizing people against these types of cancers, it would make a huge impact in the medical field by stemming the tide of cancer-related deaths.
However, a significant number of possible mutations for these cancers make it incredibly difficult to create a personalized immunization program. If scientists want T-cells to target cancerous cells, they will have to create a mutated cancer protein that tumor cells can separate into fragments to send to the cell’s surface so that T-cells will recognize and attack them.
Making that mutated protein is tricky, but possible with modern sequencing methods. The hard part is that there are thousands of possible mutations, and finding the correct one for immunization is virtually impossible for a team of people to do. Algorithms, however, don’t have the same limitations as humans, so researchers want to use them to see which parts of a mutated protein that T-cells might notice.
The research was spurred on by a 2014 study that used algorithms to cure cancer cells in mice, and since then multiple biotechnology startups have been created to explore prescriptive algorithms that can make custom vaccination options. Each laboratory has a different algorithm, and the Parker Institute and the Cancer Research Institute have launched their challenge to see which is more efficient.
The Battle Never Ends
Some researchers believe that the algorithms may never be entirely accurate at predicting the right mutation to influence T-cell responses. According to Drew Pardoll, “We don’t yet know enough about the rules to make perfect predictions. You can algorithm until the cows come home and you’re not really going to know if you’re improving things.”
Despite this opinion, even Pardoll sees merit in the idea behind the research. It is hard to doubt its potential in light of evidence that shows that imperfect prescriptive algorithms can still create an effective vaccination. So, maybe the algorithms will never be perfect, but they are a step in the right direction.
The competition between algorithms of 30 laboratories will determine which is most effective. This is a battle where cancer patients and all people alike should be keenly interested in the outcome.