My primary research work is studying the collective properties of biological systems by theoretical modeling and computer simulation based on statistical mechanics. Of particular interest are those biological issues involving randomness, diversity and correlations. I have developed bioinformatics methods to quantify the evolution of influenza and optimize the influenza vaccine efficacy.  My work on the bacterial acquired immunity (CRISPR) is the first theoretical description of bacterial systems to silence viral genes. I demonstrated that modularity can spontaneously emerge from the evolving protein networks.  By applying the biological evolution theory to the world trade network, I found that globalization would lead to increasingly large recessions and decreased rate of recovery from recession.

Bacterial acquired immunity (CRISPR)
Clustered regularly interspaced short palindromic repeats (CRISPR) have recently been shown to be a new type of anti-viral immune system in prokaryotes. I proposed a population dynamics model to explain the biological observation that the leader-proximal end of CRISPR is more diversified and the leader-distal end of CRISPR is more conserved [1]. This result is shown to be in agreement with recent experiments. Our results support that the diversity of bacteria CRISPR spacers is vital for survival. Differential equations modeling and individual-based stochastic simulation are applied in this study.

Influenza evolution and vaccine design
I developed a sequence analysis method named “Low-dimensional clustering” to predict the evolution of influenza [2]. By taking influenza protein sequences from public databases, this statistical method identifies a cluster around an incipient dominant strain before it becomes dominant. Vaccines designed using this method are shown to be higher in efficacy than currently used methods.


Modularity and hierarchy in complex networks
Modular and hierarchical structures are commonly observed in complex network. Deem’s group proposed a theory that modularity can spontaneously emergence in complex networks with the condition of experimental change and horizontal gene transfer. By studying how the protein network evolves in E. coli and yeast, I provided evidence from nature to support the growth of modularity [5].  By calculating the evolutionary rate of regulatory genes that are involved in body plan development, I found that genes in “core modules” evolve slower than those in “peripheral modules [4].” This work supports Eric Davidson’s controversial theory of hierarchical evolution of animal body plan [6]. By applying the evolution theory to world trade network systems, I found that lack of hierarchical structure in current trade leads to increasingly large recessions during economic crisis [3]. Globalization, which reduces hierarchical structure, should decrease rate of recovery from recession, in contrast to standard economic understanding.

 

---

References:

[1] J. He and M. Deem (2010), Phys. Rev. Lett. 105:128102.

[2] J. He and M. Deem (2010),  PEDS To appear.

[3] J. He and M. Deem (2010), Phys. Rev. Lett. To appear.

[4] J. He and M. Deem (2010), Developmental Biology, 337:157

[5] J. He, J. Sun and M. Deem (2009), Physics Review E. 79:031907. Also in Virtual Journal of Biological Physics Research

[6] E. Davidson and D. Erwin (2006), Science,  311:796

[7] J. Weinstein, et al., (2009), Science, 328:807