Design Issues in Implementing a Portable Sample Tracking and Analysis Research Support (STARS) System for PCR Based Microarray Research

Using a PCR (polymerase chain reaction) process to design and produce DNA microarray slides necessitates tracking thousands of samples over a long period of time. In addition, up-to-date information about the samples must be readily available when analyzing experiments done with the resulting microarray slides. A human chromosome 22 project (HC22) at Yale\´s Center for Excellence in Genomic Sciences (CEGS) used manual sample tracking and handcrafted data files for analyzing the array. The difficulties of manual sample tracking on this project limited the array layout and increased the development time. For subsequent microarray development projects at the Yale CEGS it was decided to build the comprehensive Sample Tracking and Analysis Research Support (STARS) System described herein. The system had to be portable since there are many projects at Yale and elsewhere that require automated microarray sample tracking. Informatics issues in designing and implementing CEGS/STARS are described. The system tracks plates, PCR scoring, microarray slides, and slide printings as core features. CEGS/STARS can support multiple projects and currently there are three sample tracking projects based on the same code and data base design: HC22 and HPR (human promoter) for projects in the Yale CEGS Center, and FLY for projects in the Yale School of Medicine Drosophila Center. In addition to the core tracking features, CEGS/STARS also tracks the biological experiments performed with the resulting microarrays. CEGS/STARS enhances the PCR-based array production protocol with a "cherry picking" algorithm that supports dynamic spot site decisions. Other features include: laboratory robot script generation, lab equipment file generation, automated PCR process scoring, automated links to bioinformatics data analysis tools, annotation and re-annotation of DNA fragments, links to genome databases, and publishing results to GEO. The advanced features of the system are currently being used to pro- duce a human genome promoter array, a drosophila tiling array, and a drosophila promoter array. The source code is available to interested researchers.