Here is the list of projects which I am/was involved and the relevant questions of each that I try to answer:
Computational Biology and Cancer Genomics
- How to do statistically sound large scale drug screening for combination (targeted) therapy when only outcome of single concentration combination is available?
- Given a snapshot of a cancer tumor, can we build the order in which mutations accumulated?
- What is a relation between genes, pathways, and hallmark of cancer?
- Given a set of drugs and the genetic makeup of a patient, can we exploit cancer subtypes to suggest a more promising treatment?
High Dimensional Statistics
- Having a meaningful grouping of samples, can we make a better prediction in high dimensional regime?
- How can we handle measurement errors in high dimensional regime?
Social Network Analysis
- Who is the most influential person in a social network of customers when they can change their loyalties?
- Given crowd-sourced noisy labels of tweets, can we learn to classify sentiments of a tweet?
- Assume we can modify social ties. What are the best links to add to a social network to increase the impact of a viral marketing campaign?