SEI: Machine Learning Approaches for Biological Image Informatics

基因表达空间模式的遗传分析依赖于在发育中的动物的特定发育阶段（时间）存在或不存在基因产物的直接可视化。为了有效利用快速增长的高通量果蝇胚胎基因图谱，人们正在开发一种寻找重叠表达模式基因的计算框架。第一步是开发一个学习系统，通过图像分析自动识别发育阶段。一旦确定了发育阶段，就可以比较表达模式的空间重叠。这些作品涉及高级数据类型的表示和操作，旨在使用机器学习技术构建一个新的框架。除了促进机器学习和分子生物学研究外，这项工作还将吸引学生和数据库作为教学资源。

The genetic analysis of spatial patterns of gene expression relies on the direct visualization of the presence or absence of gene products at a given developmental stage (time) of a developing animal. In order to facility effect use of the fast growing collection of fruit fly (Drosophila) embryonic gene patterns from high throughput experiments, a computational framework for finding genes with overlapping expression patterns is being developed. The first step is to develop a learning system to identify automatically the developmental stage by image analysis. Once the developmental stage is determined, expression patterns are compared for spatial overlaps. The works deals with both the representation and manipulation of advanced data types and are intended to build a novel framework using machine learning techniques. In addition to facilitating both machine learning and molecular biology research, the work will engage students and the database used as a teaching resource.