In this paper, we propose a new method for the inference of S-system models of large-scale genetic networks from the observed time-series data of gene expression patterns. The proposed method employs a technique to decompose the genetic network inference problem into several subproblems. The S-system parameters are estimated by solving these decomposed subproblems. In addition, the proposed method estimates the initial levels of the gene expression. The estimation of the initial gene expression levels is necessary when the noisy time-series data are given. We verify the effectiveness of the proposed method through the genetic network inference problems.

View full abstract

We report the molecular dynamics (MD) simulations of an asymmetric and mixed phospholipid bilayer with NaCl and cholesterols, and that with NaCl and no cholesterol. The extracellular sides of the membranes were constructed with phosphatidylcholine (POPC) molecules and the intracellular sides were phosphatidylethanolamine (POPE) and phosphatidylserine (POPS). The results showed that the cholesterol functioned to stabilize the phospholipid-mixed asymmetric membranes by making the lipid molecules more ordered and extended. In addition, it was revealed that the asymmetric feature affected the distribution of water molecules and the thickness of the membranes. The mixed phospholipid species dominated the ion distribution, such that Na⁺ ions were found more frequently on the intracellular side along the distribution of the POPS molecules. These results suggested that the asymmetric and mixed characters had a significant influence on the environment around the membrane, which would also affect the membrane proteins.

View full abstract