Research Interests

How does a genome specify the properties of an organism?

To answer this question, it is essential to understand how genes interact with each other to carry out a biological process. My research is aimed at achieving a network view of genetic interactions. In particular, I am interested in developing high-throughput methods to conduct genome-wide studies of genetic interactions.

My research pipeline is composed of two steps: hypothesis generation and experimental verification. To reach high-throughput efficiency, I apply bioinformatic tools to extract hypotheses and automation systems to facilitate experiments.

• Computational prediction of genetic interactions

The wealth of genomic information provides an unprecedented opportunity for new computational strategies to infer genetic interactions. By combining various data sets (such as phenotype, expression, interactome) from multiple species, and applying proper statistical models, we can predict genetic interactions with high accuracy and provide significant clues for biologists. We have applied this strategy to identify functionally interacting genes in the metazoan C. elegans using data from three species (worm, fly, yeast). (for details, see Publications). We are in the process of expanding the system to include data from more species and extend the predictions to genetic interactions in more organisms. We are also interested in exploring different techniques and algorithms for such data integration.

• Quantitative genetics for experimental testing

To explain the value of experimental verification, here is a quote from Henry IV by Shakespeare.

Glendower: I can call spirits from the vasty deep.
Hotspur: Why, so can I, or so can any man; But will they come when you do call for them?

OK, everyone can make predictions. What matters is "does it work".

To experimentally verify a genetic interaction, we will inactivate two genes simultaneously and screen for suppression or enhancement of a phenotype. Unlike conventional screens that focus on qualitative differences such as synthetic phenotypes, we seek to examine genetic interactions with quantitative analysis. We want to derive quantitative mathematical models from data on known interactions, collect data on new gene pairs at a genome-scale, and apply the model to detect new interactions.

To handle the labor-intensive nature of quantitative analyses, we employ automation in our experiments. We have a prototype of a computer vision system that measures worm body lengths. We are exploring a number of technologies (e.g. robotics, microfluidic devices) to improve the performance of the system and to extend its application to analyze other morphological phenotypes and other organisms.

I have a dream that one day robots will do all the experiments for me.

Weiwei Zhong
Last modified: August 7, 2006