Development Of Web-based QTL Mapping System

Most of the biological, medial and agricultural important traits are complex traits that result from the segregation of alleles at multiple quantitative trait loci (QTLs) with effects sensitive to genetic, sexual, and environmental factors. Yang et al. (2007) proposed a full-QTL model to explore the genetic architecture of complex trait in multiple environments, which includes the genetic effects of multiple QTLs, epistatic effects, their interaction effects with environments. An F-statistic based on Henderson method III was used for hypothesis tests. In each of the mapping procedures, permutation testing was exploited to control for genome-wide false positive rate, andmodel selection was used to reduce ghost peaks in F-statistic profile. Parameters ofthe full-QTL model were estimated using a Bayesian method via Gibbs sampling.This method can be applied to map QTL for most of experimental populations, and can improve the precision and efficiency of QTL detection.In the present study, a web-based QTL mapping system was developed utilizing Java and JSP (Java Server Pages) to apply the aforementioned QTL mapping method. The B/S (Browser/Server) model was employed for connecting the user interface (webpage) and server. User can access the URL http://ibi.zju.edu.cn/software/qtlnet work/webservise to submit the data files and set mapping parameters for remote computation. Once the server completes the data analysis, the results will be saved in text file and sent back to the user by email. User can open the result file by any text editor, or by QTLNetwork software to visualize the mapping results.By developing the web-based QTL mapping system, we studied and applied the Java and JSP techniques, and designed the JSP solution for QTL mapping. Since the system was developed based B/S model, it can be accessed by user using any operating system. As an extension of QTLNetwork software, this system provided a new tool for QTL analysis of complex traits. Especially, for large dataset, the computing time can be largely reduced by running the data on server with high-computational performance, and it is not necessary to have online waiting during the running process. The proposed solution for web-based QTL analysis can be easily modified for some other applications of complex trait analysis, which provided a reference example for design of biostatistics software.