Nikhil M
- Research Program Mentor

PhD candidate at Stanford University

Expertise

Machine Learning, Statistical Learning, Statistical Genetics, Functional Genomics

Bio

I am interested in the field of statistical genetics, where I combine my passions of statistics, computer science, and genetics to better understand human traits and diseases. I love to understand and apply statistical and machine learning models. I have trained at the University of Cambridge as a Master's student, where I worked at the Wellcome Sanger Institute for a year to better understand the genetics underlying sepsis. Now, I have started a PhD at Stanford University in Genetics, where I also plan to pursue a Master's in Statistics! When I am not conducting research, I love to play musical instruments! I am a violinist by training, but have picked up the piano and guitar along the way. I also love to hike - the national parks in California provide beautiful places to do so!

Project ideas

Project ideas are meant to help inspire student thinking about their own project. Students are in the driver seat of their research and are free to use any or none of the ideas shared by their mentors.

Machine learning approaches to better understand disease

The field of genomics is very good at sharing data publicly! Large studies conducted on humans have accrued millions of data points across many publicly-available data sets. These provide an amazing substrate to test the utility of machine learning approaches to better understand disease. A machine learning approach that is good at predicting disease from such data can be very helpful for medical professionals when making diagnoses. In this project, an interested student may start by downloading data from a few studies of interest and using machine learning or statistical learning approaches to identify genes that may contribute to disease. Students can compare various approaches and identify methods that work well in genomic datasets.

What can we learn from classical music?

Classical music spans multiple centuries, each with their own unique styles. This can make it quite fun to ask some questions using machine learning and statistical approaches: (1) Can we find out who composed a piece by looking at it? By hearing it? (2) What makes a composer famous? Is there anything similar about music across centuries? (3) Can computers make music that is as good as a trained composer?

Coding skills

Python, R, Stan

Languages I know

Hindi and Marathi, native

Teaching experience

Starting early in high school, I was the president of both the guitar club and the computer science club. In this position, I taught middle and high school students how to play the guitar and how to code. I fell in love with mentorship, and have continued to teach ever since. Later in high school, I taught other students in academic seminars. As an undergraduate, I taught two elementary school students to play the violin for four years. I was also a mentor to junior undergraduates that joined my research laboratory.