Here’s our first interview of 2022 !
What do you do when you want to explore biology, but from a mathematical angle? Well !, you become an mathematical biologist !
Ananthu James (PhD), our next pathbreaker, Epidemiologist, conducts data centric research (in collaboration with an international public health network) that can directly influence clinical guidelines or policies concerning health/infectious diseases.
Ananthu talks to Shyam Krishnamurthy from The Interview Portal about the huge potential in applying his quantitative background in Physics, Mathematics and Biology to address pressing challenges in evolution and human/animal/plant health.
For students, the world is your oyster; Biology and Mathematics are not mutually exclusive. Kick off the new year with a bang !!!
Ananthu, Your background?
I am from Kerala, and my parents are retired government employees. I studied in Kerala until the completion of my bachelors in physics. Interestingly, my career or career interests were shaped by some of my non-academic activities in my early days. In my early childhood, I used to like spending time in my grandfather’s farmland. I used to love feeding (stray) cats at home, and they became residents of my compound for around a decade. I always enjoyed spending time with them. Moreover, I actively watched many wildlife documentaries on Animal Planet and National Geographic. These experiences made me fall in love with the living world. Besides, I followed cricket and football on television, and apart from the game, my interest was also the scoreline (I had a weird interest in numbers/data, which might have helped me in moving to data science later in my career).
What did you do for graduation/post graduation?
I did my bachelors in physics from the Mahatma Gandhi state university in Kerala. Following that, I cleared IIT JAM, and did masters in physics from IIT Madras. I also did my PhD (Mathematical biology -Population genetics) from Jawaharlal Nehru Centre for Advanced Scientific Research, Banaglore.
What influenced you to pursue such an offbeat, unconventional and uncommon career?
My parents encouraged me to go for basic sciences, preferably physics, instead of engineering (although they expected me to settle with a teaching job following my masters or PhD). And, despite having a slightly higher affinity for biology, I ended up following the general trend of choosing physics among basic science subjects. During my first year of BSc physics, I got selected for the KVPY fellowship following a national level exam and interview. This was a turning point in my career, which gave me the opportunity to do research as part of summer projects. Thanks to my first mentor Dr. Vincent Mathew who gave me a lot of freedom, I enjoyed research and remained positive about an academic career. This also motivated me to continue pursuing physics for my masters. However, at the time of joining my masters and even my PhD, other than having interest in the respective subjects, I did not have much of an idea about my future plans. PhD was more like an adventure for me. Later, thanks to the exposure I had from three national institutes, I realized that while combining my profession with personal interests (so that my happiness would not be compromised), I should also be part of some impactful work.
How did you plan the steps to get into the career you wanted? Or how did you make a transition to a new career? Tell us about your career path
After pursuing physics until my masters, owing to my childhood interest in life sciences, I wanted to explore the possibilities of pursuing biology for PhD. At the same time, I liked the mathematical/quantitative aspects, and I never wanted to completely stay away from that. This is how I decided to join a lab that works on mathematical biology at the JNCASR, Bangalore, among the options I had at that time. Initially I found the work very cool, since it involved deriving new mathematical expressions from complex equations that modeled biological evolution.
My PhD was centered around mutations (genetic changes) in asexual populations (organisms in which reproduction is simply passing the genetic material (like RNA or DNA), without the need of another individual), such as some of the microbes. Even among asexuals, I considered haploid organisms, in which there is only one single “copy” of the genetic sequence. (Humans and other advanced organisms have two such copies, and are called diploid.) In asexuals, mutations are the only source of variation. In an environment that has not been changing for generations, owing to evolutionary forces, asexual populations eventually become well-suited or “adapted”. Thus, any existing organism also is likely to be well-adapted. Therefore, any genetic change (mutation) is not welcome and will only decrease the survival chances of the individual. Because of this, in asexual populations adapted to an environment, the mutation rate (the number of mutations per generation) is prone to decrease. (Interestingly, there are also mutations that cause change in mutation rates!) The purpose of my PhD was to study the factors that would depend on mutation rate reduction in asexual populations already adapted to an environment. Some of these factors were the effect of each mutation on survival chances of an organism, the magnitude of change in mutation rate, the presence of genetic interactions, etc.
This process can be modelled using simple recurrence relations connecting parental and offspring generations. To incorporate the complex effects of mutations on survival chances of an individual, I used some mathematical functions. While the resulting equations can be solved computationally, I also obtained simple but intuitive closed form expressions using some assumptions and simplifications. Independently, I performed stochastic simulations to capture this phenomenon and match my simple mathematical equations. The mathematical areas involved in my work were probability theory (all my published papers have the term ‘probability’ in common), stochastic processes, series, special functions, linear algebra, and calculus.
In 5-6 years, I completed my PhD with three publications, including a single author paper in the Journal of Theoretical Biology.
Following my PhD, with the aim of getting exposed to other biology-related areas, before deciding my next major step, I joined as a PostDoc at the Chemical Engineering Department, IISc Bangalore, with Prof. Narendra Dixit who addresses problems on infectious diseases and immunology using mathematical modeling. I wanted to do something that would have an overlap with evolution, which I pursued during my PhD.
For my PostDoc, my primary focus was on analyzing HIV spread, its transmission, and the associated factors. Initially, I employed stochastic simulations to study this. However, ever since I started working on my main problem (described in detail at a later stage here), I had to rely on statistics, and I did the following two things mainly – test for pairwise comparisons and regression. The former involves comparison of two different groups. An example would be comparing the viral loads (viral copies per blood) in HIV infected men and women. Apart from sex, these two groups may differ in various other aspects like age and ethnicity. Hence, pairwise comparisons cannot reveal the individual effect of only a particular factor. This is where a regression analysis is required. Regression helps in estimating the individual effect of sex, age, ethnicity (and other factors) on viral load, and clearly tells us whether men or women are more likely to have high viral loads when all other factors are the same between them.
Other than HIV, I also did research on COVID-19, and that also involved statistics. This work also had pairwise comparisons and regression.
Although my initial plan was to consider this postdoc as a short period in my career, I found a positive environment and the research questions truly impactful. Moreover, I got time and peace to learn new aspects and develop my interests. Hence, I decided to spend considerable time in this group. Besides, I saw my colleagues happily transitioning to the industry and remaining there, a trend which was rare in basic science departments. Until then, I had the false notion that anything other than academia cannot make people happy.
After spending around three years at IISc, I wanted to consider other options, especially going abroad. But, COVID-19 spoiled my plans. This is when I came across a recently founded international public health network (the GRAPH Network), which in collaboration with the World Health Organization (WHO), focuses on capacity building in Africa, and addresses problems concerning global health. Being able to appreciate their mission, I started working part-time with them initially on their academic publications, which substantially involved data/statistics. (I acknowledge my postdoc supervisor Narendra again, for being flexible enough to let me explore other career opportunities like this, while keeping me as a full-time postdoc fellow in his lab.) At the GRAPH Network, I found collaborators from diverse disciplines (and almost all continents), including mathematics, medicine, epidemiology, ecology, and social sciences, with many of them having excellent programming skills and many more years of experience than me. My strong quantitative background (masters in physics and PhD in mathematical biology) became very handy in grasping the necessary statistical aspects quickly, though I fully thank my supportive colleagues from whom I learnt many of these. This collaboration has been extremely successful, and also resulted in three paper submissions so far. Later, I also got an official contract from WHO in order to support the epidemiological data analysis in the African region.
Can you explain the concept of epidemiological data analysis ?
A simple definition of epidemiology is that it is the study of the main determinants of a disease. The example I mentioned above while explaining my HIV work also involves epidemiological data analysis, since it deals with identifying what decides high or low viral load in individuals – whether it’s sex, age, ethnicity, or any other factor. Thus, epidemiological data analysis reveals important information regarding the spread or severity associated with diseases or health events. For example, once we know that a particular infection causes more deaths among children, we can immediately take the necessary preventive measures (such as school closures etc.) to ensure that children remain unexposed to the pathogen. Or, if a Pharma company knows that a particular medicine manufactured by them causes adverse effects for those with high blood pressure, they can specifically state that the medicine must not be prescribed for such individuals. In this way, epidemiological data analyses often have immediate impact.
How did you get your first break?
I wanted a small break following my PhD thesis submission to decide the next step carefully rather than jump into something without proper thinking. I spent some time (~3 months) traveling to the Himalayas and the Western Ghats. Meanwhile, I became more positive about trying out new areas, and contacted Prof. Narendra Dixit (who is well known to those who work in theoretical/computational biology, especially in Bangalore), who offered me a postdoctoral position in his lab.
What were some of the challenges you faced? How did you address them?
Challenge 1: Unfortunately, my initial excitement towards my PhD did not last long. With time, I realized that the models I was using were based on too many assumptions, often far from reality. Overall, despite my PhD being academically successful, by the time I completed it, I was not able to appreciate it fully. A big lesson from this experience was that whenever possible, mathematical models should be driven by the latest data from experiments or real life scenarios (although there are situations where simple models are preferred). I even questioned the whole point of remaining in academia because I wanted more personal satisfaction from my work. I wanted to do something that would have a more direct impact on the world. Sadly, I was neither well aware nor positive about industry or other non-academic possibilities at that time. Ironically, I did not have the breadth or skillset to directly make a smooth transition to other fields even within academia. Despite all that, I decided to make a career change to infectious diseases epidemiology. Although things took time, my decision turned out to be good, also thanks to my supportive postdoc lab and colleagues from GRAPH Network.
Challenge 2: The COVID-19 pandemic completely changed my plans – concerning my career as well as personal life. Traveling (particularly to forests or hill stations) was a big source of entertainment for me, which used to get me recharged for my work. I had to stop all sorts of traveling and confine myself to just my home for many months – a fate not uncommon to others during the pandemic. Although I was able to grow academically, my mental health kept deteriorating. In between, I had also joined the GRAPH Network, and I had to work on multiple tasks. However, I found absolutely no motivation to do anything. As a last resort to get out of the situation, I went in for a big commitment at that time – adopting a stray puppy from a local shelter. Since I always enjoyed the company of animals, I was sure that a companion animal would keep me happy and entertained. This proved more than successful, and I became very productive ever since the existence of the dog in my life. This is when I learnt the importance of having a balance in life. Mental health is a must before we plan anything else.
Challenge 3: I had never done multitasking before joining the GRAPH Network. There were times when I had to do three-four different things at a time, due to the work at IISc and the GRAPH Network. However, since I had reasonable experience in the field (infectious disease epidemiology), and there was a decent overlap between various projects I was working on, things were easier than what I expected. In fact, I was also able to manage the work at these two places synergistically. I also learnt how to effectively prioritize tasks.
Where do you work now? What problems do you solve?
I work at three places now. I have a full time position as a postdoc with Prof. Narendra Dixit, IISc. I have been working part-time with the GRAPH Network for more than a year. Recently, I also got a part-time contract with the World Health Organization (WHO), which is my third workplace currently.
The commonality across all my three workplaces is infectious disease epidemiology and the R programming language. The difference is the following. My work at IISc is purely research involving HIV and COVID-19, driven by immunological or evolutionary hypotheses, with data science/statistics employed to support them. My work with GRAPH Network so far has also been mainly research, but without necessarily having hypotheses. This work also is extensively supported by data/statistics. I mostly use R for these purposes. On the other hand, with WHO, I am involved in writing scripts in R for extracting and visualizing daily/weekly data from Africa pertaining to the COVID-19 pandemic.
A typical day involves coding in R, emails, phone calls, attending meetings, and working on publications.
I like working with data, and I am happy that it has direct practical implications (mainly because the work is concerning infectious diseases). Moreover, I work with a set of people who are passionate and, at the same time, flexible about the work.
How does your work benefit society?
The research I’m doing can directly influence clinical guidelines or policies concerning health/infectious diseases. Apart from research, the work with GRAPH Network will have many other exciting aspects in the near future, such as developing a platform to train researchers etc. My work with WHO helps in assessing the pandemic situation in each African country so that WHO can alert the concerned countries or take the appropriate measures.
Tell us an example of a specific memorable work you did that is very close to you!
While examining the data for a side-project at my postdoc lab, I noticed a clear difference in both the immune cell counts and drug resistance mutations between HIV infected homosexual men and heterosexuals. (I had already mentioned my interest in numbers/data, due to which I followed sports, especially, the score lines.) I suspected some evolutionary forces to be responsible for the above differences, and discussed it with my Professor. He encouraged me to take this up as my main problem and solve it. For this purpose, I started learning statistics, and this paved the way to my entry into data science. Solving this problem was a very long journey, and I had to carefully analyze various evidences I obtained concerning the social situation and behavioral patterns of homosexuals and heterosexuals in various regions as well as the already known roles of age, sex, viral load, HIV genetic composition (also known as subtype) etc. on immune cell count. Further, this led me to understand and develop interest in epidemiology and global health. Eventually, I also had to rule out an initial hypothesis I had regarding this observation, clearly demonstrating the complexity of biological systems and how careful we have to be regarding our assumptions. In terms of the learning experience, the impact and duration of the work, and being responsible for a change in my career trajectory from mathematical biology to epidemiology, this has been the most memorable work.
Another memorable work is the single author paper I published during my PhD. That boosted up my confidence substantially, and played a crucial role in continuing my academic career, but at the same giving enough courage to consider working on other areas.
Your advice to students based on your experience?
With a strong mathematical background, one can excel in so many fields, provided they are genuinely interested in the new field to understand it in the right sense (without blindly buying any mathematical model). Interdisciplinary research is becoming increasingly common.
Whenever we realize that we are not in the right field, it’s better to consider shifting. The process can take time, but it’s worth doing. For that, it’s important to be aware of other fields so that we can apply our expertise. Breadth and depth are crucial for a successful career.
Most of the work we do now is collaborative. Through collaborations, we can quickly acquire new skill sets, and be involved in tasks which we cannot independently perform. A very under-rated skill is certainly the ability to smoothly work with others. Therefore, even for career or personal growth, I always recommend work environments where employers and co-workers are empathetic.
Basic programming skills involving R and Python (hopefully these two will not become outdated by the time somebody reads this article!) are relatively easy to acquire, and they are very powerful. Moreover, ability and experience in working with data is very useful.
My personal experience is that along with the outcome, it’s important to be able to enjoy the process we are involved in.
I am currently happy with the work I do, combining data science and health. I am able to appreciate its impact. However, I may like working on related fields such as animal/plant health as well as the possibility of involving evolutionary aspects.
Learning data science has been an excellent step in my career. There has been a big rise in my potential future possibilities in the last one year, thanks to data science (and also the collaborations I have been part of). I therefore want to continue mastering this area, even independently of health/biology. Aided by my strong quantitative background, I feel confident of contributing to various fields (if I find interesting possibilities there).