Project during postdoc:
1. Understanding multi-trait GWASs toward a more effective usage.
Despite the rich human genetics data that researchers have successfully collected, there are still many undetected SNPs that are affects human traits including diseases ("missing heritability problem"). Method development is one direction to tackle this challenge and there have been several advanced statistical methods to detect associated genetic variants. Given multiple methods, we need to know which methods work better under which conditions. To address this, I'm working on GWAS methods, including classical and multi-trait GWASs, to compare and use them more effectively.
2. Network-based data integration toward a better link between genetics and traits.
Another challenge to overcome the missing heritability is to integrate multi-layered biological data. I am working to integrate gene-gene networks and GWAS results to better understand links between human genetics and some traits (e.g. asthma) by integrating multi-layered data and using methods such as machine learning and network-based methods.
Project during PhD program:
1. Relating explicit spatial structure with biodiversity in a metacommunity
I developed a model of metacommunity with explicit spatial structure (network) with both stochastic and deterministic processes.
In this model, stochastic processes are such as dispersal and drift and affect all species in the same way, whereas deterministic processes are ones that are associated with fitness difference and environmental difference and affect different species differently.
With this model, I was able to link better the spatial structure of meta-community and biodiversity patterns and stability of the system.
2. Network analysis of larval dispersal connectivity in marine systems
I applied network analysis to multiple larval dispersal connectivity networks, to find out characteristic network features in marine networks and link to the mechanisms understood by the above project. I used a few marine networks that my collaborators (Dr. Satoshi Mitarai's unit in OIST) shared with me, which are all actually models of connectivity networks developed with ocean current simulations.
