The world's human population is
expected to grow
past 9.7*10^9 by 2050
, and our
global temperature is expected to increase by at least
in the same time frame. I'm interested in technologies that
prepare us for climate change and the demands of a growing
Our agricultural production is expected to fall short of global demand in the coming years as the global population swells and living standards rise. Agricultural technology must keep pace if we are to sustain 10B people.
I focus my research on key areas of biology that stand to benefit from the development of new statistical theory and computatinal methodology. Broadly, I'm working on statistical methods for genomics and computational methods for climate models.
I am interested in new methods of computational genomics and how they inform our understanding of the molecular and regulatory mechanisms underlying flowering time. Flowering is a key agronomic stage in plant development linked directly to how much food a crop produces. Though broad phylogenetic comparisons of flowering regulatory networks in wild and domesticated crops are underway, this information has yet to be incorporated into agricultural portfolio models that account for climactic uncertianty. Integrating flowering regulatory networks into decision support systems allows farmers and researchers alike to better plan for bad weather .
Following the central dogma of molecular biology, genetic
expression drives phenotypic development. Knowing which
genes determine which phenotypes is vital for
understanding cancer, crop growth, and synthetic organisms.
To build this knowledge, a variety of relatively successful
methods have been developed
Among these methods are traditional Elastic-Net-like restricted generalized linear models, graphical methods, and attempts to incorporate prior biological knowledge. Yet linking phenotypes to genotypes and expression profiles remains an open challenge, and big data is being made available for open exploration. . New methods utilizing infor mation theory and topological data analysis are emerging as new, promising areas of statistical insight in functional genomics.
How do we use genomic, sensor, and high throughput phenotype data to build better models and predictions for biological systems? Recent advances in computing, mathematics, molecular and systems biology are enabling the kind of cross-scale, cross-discipline synergy these insights need to emerge. I take inspiration from the Runcie Lab, the Edwards Lab, and the MIT climate modeling group.
Here at Davis, I'm building collaborations between plant science and statistics to understand where our crops will grow in increasingly unstable environments. Currently, I'm working on topological methods for genomic data analysis for flowering pathways in rice. Rice feeds much of the world and is a model short-day crop, making it an important and tractable first step. Nature's special issue on rice is a great read.
Peoples' genomes, their locations, and conversations are all private. This privacy must be respected by practitioners, enforced by government, and watched diligently by third parties. A good primer can be found here. In addition to privacy concerns, what is considerably more pressing is the proper use and democratization of new data technologies. We're not building nukes, but these tools are still powerful and we must be diligently responsible scientists.