A mesh generation and machine learning framework for Drosophila gene expression pattern image analysis

Abstract

Background: Multicellular organisms consist of cells of many different types that are established during development. Each type of cell is characterized by the unique combination of expressed gene products as a result of spatiotemporal gene regulation. Currently, a fundamental challenge in regulatory biology is to elucidate the gene expression controls that generate the complex body plans during development. Recent advances in high-throughput biotechnologies have generated spatiotemporal expression patterns for thousands of genes in the model organism fruit fly Drosophila melanogaster. Existing qualitative methods enhanced by a quantitative analysis based on computational tools we present in this paper would provide promising ways for addressing key scientific questions.

Results: We develop a set of computational methods and open source tools for identifying co-expressed embryonic domains and the associated genes simultaneously. To map the expression patterns of many genes into the same coordinate space and account for the embryonic shape variations, we develop a mesh generation method to deform a meshed generic ellipse to each individual embryo. We then develop a co-clustering formulation to cluster the genes and the mesh elements, thereby identifying co-expressed embryonic domains and the associated genes simultaneously. Experimental results indicate that the gene and mesh co-clusters can be correlated to key developmental events during the stages of embryogenesis we study. The open source software tool has been made available at http://compbio.cs.odu.edu/fly/.

Conclusions: Our mesh generation and machine learning methods and tools improve upon the flexibility, ease-of-use and accuracy of existing methods.

Figure 1

Left: Subdivision of an ellipse (pink) based on equal radial angles (dashed black lines) leads to inaccurate boundary interpolation (blue). Center: A more accurate subdivision (solid black lines) based on equal lengths of interpolating segments. Right: Euclidean projection (q) from a point (p) on the ellipse (green) onto the boundary of the embryo (red) is more accurate than a projection along a radial line (r).

Figure 2

Clusters of mesh elements when the number of clusters is varied from 10 to 30 with a step size of 5 (left to right, top to bottom) on stage 4-6 expression patterns. The first figure in the first row shows the fate map of the blastoderm [27].

Figure 3

Clusters of mesh elements when the number of co-clusters is set to 39 as in [8], and when the number of mesh elements are set to 300, 600, and 1000 (left to right). In these figures, colors are used to visualize clusters so that mesh elements in the same cluster are in the same color, and those in different clusters are in different colors.

Figure 4

Comparison of the total numbers of enriched gene ontology terms obtained from our co-clustering method and the affinity propagation method used in [8]. The reported numbers here are the total number of terms in each cluster.

Figure 5

Significance levels of enriched GO biological process terms on the data set of 2693 images. Scales indicate significance levels, and only terms with ≥90 significance are shown. A total of 39 co-clusters have been generated, and only co-clusters with enriched terms are shown. The corresponding mesh clusters are shown in Figure 6.

Figure 6

Mesh clusters corresponding to the gene clusters in Figure 5. To establish a correspondence between this figure and Figure 5, each mesh element is labeled with its cluster membership, and the corresponding gene clusters are shown as columns in Figure 5.