Heterosis, or hybrid vigor, refers to the superiority of F₁hybrid performance over the mean of its parents (mid-parent heterosis) theoretically, or the performance of better parents. It has been discovered in many species of plants and animals as well as in humans, and played an important role in enhanced agricultural production, especially in maize, rice and sorghum although the mechanism have not been elucidated. We studied the molecular basis of heterosis with a combined genomics and systems biology approach using model organism maize. We profiled the expression of 39 genes that were most differentially expressed (DG) between the mid-parents and their F1 hybrid (Mo17 x B73) in the 13V-satged, developed whole ear shoots of 13 inbred lines and their 22 F1 hybrids grown in the field trails and phenotyped their 13 traits significant for grain yield. The results showed that gene expression varies significantly among inbreds, among hybrids and in heterosis. The gene clustering heat map and gene action networks in inbreds and hybrids were constructed respectively based on their gene expression profile. According to these pattern analyses, we find dramatically difference between inbreds and their hybrids, although the differential expression varies across different hybrids. Our results also suggest that gene networks are altered from inbreds to hybrids, including their gene contents and wire structures. Last but not least, we have determined the genetic variation correlations between the gene expression and trait performance and constructed the gene networks for the development of 12 of the 13 traits that varied significantly among genotypes. This has led to identification of genes significantly contributing to the performances of the traits, with 1 – 16 genes per trait. These results have indicated that heterosis results not only from altered expression level of corresponding genes between inbreds and their hybrids, importantly, also from the altered gene action networks and expression patterns. These alternations could be derived from gene actions in a manner of additivity, dominance, over dominance, pseudo-overdominance, epistasis and/or their combinations. Therefore, our findings provide a better understanding of the underlying molecular basis of heterosis. The genes identified for the traits will provide tools for advanced studies of the trait heterosis and could be used as tools for their heterosis breeding in maize. The strategy developed in this study will provide an effective tool for studies of other complicated, quantitative traits in maize and other species.

The formation and persistence of species are the subject of much debate among biologists. Many species of Drosophila are behaviourally isolated, meaning that heterospecific individuals are not attracted to one another and do not interbreed. Often, this behavioural isolation is at least in part due to differences in pheromonal preference. Drosophila pheromones are long-chain cuticular hydrocarbons (CHCs). Desaturases are enzymes that are important for the production of CHCs. This thesis investigates the evolution of the gene family across 12 species of Drosophila. Desaturase genes were located in all species. Some genes, those that have previously been shown to have important roles in pheromonal communication, have experienced duplication and loss in several species. Two previously undiscovered duplicates were identified. Generally the desaturase gene family is governed by purifying selection, although following duplication these constraints are relaxed and in some cases duplicated genes show compelling evidence of positive selection. One of the loci under positive selection, the novel duplicate desat1b of the obscura group, was found to have a sex-biased expression pattern and alternative splicing in its 5′ UTR. In RNAi knock-down experiments of desaturase gene function in D. melanogaster, several desaturases were shown to affect CHC profiles of males and females, including some that were previously unlinked to CHC production.

No description available.

In this thesis, the expression variation problem in two-dimensional (2D) and three-dimensional (3D) face recognition is tackled. While discriminant analysis (DA) methods are effective solutions for recognizing expression-variant 2D face images, they are not directly applicable when only a single sample image per subject is available. This problem is addressed in this thesis by introducing expression subspaces which can be used for synthesizing new expression images from subjects with only one sample image. It is proposed that by augmenting a generic training set with the gallery and their synthesized new expression images, and then training DA methods using this new set, the face recognition performance can be significantly improved. An important advantage of the proposed method is its simplicity; the expression of an image is transformed simply by projecting it into another subspace. The above proposed solution can also be used in general pattern recognition applications. The above method can also be used in 3D face recognition where expression variation is a more serious issue. However, DA methods cannot be readily applied to 3D faces because of the lack of a proper alignment method for 3D faces. To solve this issue, a method is proposed for sampling the points of the face that correspond to the same facial features across all faces, denoted as the closest-normal points (CNPs). It is shown that the performance of the linear discriminant analysis (LDA) method, applied to such an aligned representation of 3D faces, is significantly better than the performance of the state-of-the-art methods which, rely on one-by-one registration of the probe faces to every gallery face. Furthermore, as an important finding, it is shown that the surface normal vectors of the face provide a higher level of discriminatory information rather than the coordinates of the points. In addition, the expression subspace approach is used for the recognition of 3D faces from single sample. By constructing expression subspaces from the surface normal vectors at the CNPs, the surface normal vectors of a 3D face with single sample can be synthesized under other expressions. As a result, by improving the estimation of the within-class scatter matrix using the synthesized samples, a significant improvement in the recognition performance is achieved.

In this thesis, the expression variation problem in two-dimensional (2D) and three-dimensional (3D) face recognition is tackled. While discriminant analysis (DA) methods are effective solutions for recognizing expression-variant 2D face images, they are not directly applicable when only a single sample image per subject is available. This problem is addressed in this thesis by introducing expression subspaces which can be used for synthesizing new expression images from subjects with only one sample image. It is proposed that by augmenting a generic training set with the gallery and their synthesized new expression images, and then training DA methods using this new set, the face recognition performance can be significantly improved. An important advantage of the proposed method is its simplicity; the expression of an image is transformed simply by projecting it into another subspace. The above proposed solution can also be used in general pattern recognition applications. The above method can also be used in 3D face recognition where expression variation is a more serious issue. However, DA methods cannot be readily applied to 3D faces because of the lack of a proper alignment method for 3D faces. To solve this issue, a method is proposed for sampling the points of the face that correspond to the same facial features across all faces, denoted as the closest-normal points (CNPs). It is shown that the performance of the linear discriminant analysis (LDA) method, applied to such an aligned representation of 3D faces, is significantly better than the performance of the state-of-the-art methods which, rely on one-by-one registration of the probe faces to every gallery face. Furthermore, as an important finding, it is shown that the surface normal vectors of the face provide a higher level of discriminatory information rather than the coordinates of the points. In addition, the expression subspace approach is used for the recognition of 3D faces from single sample. By constructing expression subspaces from the surface normal vectors at the CNPs, the surface normal vectors of a 3D face with single sample can be synthesized under other expressions. As a result, by improving the estimation of the within-class scatter matrix using the synthesized samples, a significant improvement in the recognition performance is achieved.