There’s still no cure for AIDS, but scientists are using a mathematical model that might help them design more effective drugs and develop a vaccine for AIDS.
There are 100 billion new particles of the virus produced every day, Discover blogger Ed Yong wrote. At this rate, it’s hard to design a vaccine against HIV because the virus mutates so fast. However, some parts of the virus don’t mutate as much as other regions. This is the part where scientists want to focus on.
Scientists used mathematical models typically used by Wall Street traders to analyze stocks. MIT researcher Arup Chakraborty spoke to SmartPlanet about how he’s using concepts from physics, math and engineering to understand how HIV mutates.
In the interview below, Chakraborty explains how he used random matrix theory to find hidden correlations within a large set of data.
SmartPlanet: How did you use math to crack the mystery behind HIV replication?
AC: First, we have made a step forward. We have not solved the whole problem. Once confronted with a pathogen such as a virus, our immune system develops a memory of this episode.
This immunological memory enables us to mount rapid and vigorous responses upon re-infection by the same pathogen.
SmartPlanet: Okay, why is it so hard to produce a vaccine?
AC: A vaccine aims to prime the immune response to develop memory for the pathogen against which one wishes to protect people.
It’s difficult to design a vaccine against HIV because the virus mutates very rapidly. Thus, it can evade the memory immune response that is primed by vaccination.
SmartPlanet: How can you use the immune memory to stop HIV from replicating?
AC: Researchers target individual amino acids in HIV proteins that do not mutate much, so that immune pressure applied at these amino acids would promote the emergence of a mutation there. But this kind of mutation is likely to hurt the virus’ ability to replicate. Unfortunately, a second mutation can arise, that together with the first one, can restore the virus’ fitness.
SmartPlanet: What did your study involve?
We looked for groups of amino acids, not single ones, that co-evolve. For example, we looked at some combination of how mutations within a group influenced virus fitness.
SmartPlanet: How did you use math to analyze HIV sequences?
AC: We used a statistical method to analyze available HIV sequences to search for groups that may be a good target for the T cell arm of a vaccine. Based on mathematical analysis, we understand why the group is vulnerable based on the structure of HIV. We are testing some of our predictions against human clinical data.
For example, people who control HIV without medications, disproportionately target the vulnerable region we have identified. Further work is required to obtain other such regions of vulnerability, including those that could be targeted by another key arm of a potential vaccine.
SmartPlanet: How can you ever stop the HIV virus from replicating if it mutates?
AC: We think the trick is to hit it with immune pressure in such a way that it either makes multiple mutations that hurt its ability to replicate and function or it lives with immune pressure that can control the virus by killing it.
Vaccines containing the vulnerable regions only are expected to train the immune system to attack these regions. We plan on using animal models to prove this idea.
SmartPlanet: How do you see math being more integrated into the study of viruses in the future?
AC: I think that some major breakthroughs in medicine and biology will come by bringing together approaches from the physical, engineering, and life sciences with clinical studies. Indeed, the Ragon Institute of MGH, MIT and Harvard was formed with this premise to make discoveries that can create a vaccine against HIV and other diseases.
SmartPlanet: What exactly is random matrix theory? How complicated is it? What are its limitations?
AC: It is a method originally developed in nuclear physics, which aims to tease out real correlations between events. In our case, [we wanted to know] mutations in HIV from noise and random events. It has been applied in many realms of physics, used to analyze stock market fluctuations and used to determine groups of co-evolving amino acids in an enzyme.
I thought these studies might be useful for the HIV problem. It has numerous limitations, but we are trying to overcome them in future studies.
SmartPlanet: What problem are you trying to fix? Do you spend a lot of time thinking about how to solve complex problems with math?
AC: I have spent the last 11 years bringing approaches from physics and engineering to confront major basic questions in immunology. Two years ago, Bruce Walker, my clinician collaborator on this paper as well as the one in Nature last year got me interested in working on HIV. I became a founding member of The Ragon Institute.
I was very reluctant at first. But Bruce took me with him to visit the HIV hospitals in South Africa. The pain and suffering I witnessed during that visit and subsequent ones made me want to contribute what little I could to help confront this global challenge.
If you want to read about my transformation to work in immunology with the specific goal of bringing mathematical, physics and engineering approaches to the subject, there is a story in The Scientist in 2010, and another in Science News last year that you could read.
hat tip to New Math in HIV Fight [Wall Street Journal]
Related on SmartPlanet: