Site Navigation:

YMD > Current Status

(3/9/2003 summary prepared by Kei Cheung): YMD is currently being beta-tested by a number of laboratories involving microarray experiments with different organisms. These include Dr. White’s lab (Drosophila), Dr. Reinke’s lab (C. elegans), the Center of Excellence for Genomic Science Microarray Facility (Human and Yeast), and the Keck DNA Microarray Resource at Yale (Human, Mouse, Arapidopsis, etc). YMD is currently MIAME-supportive (its current design is based on a subset of MIAME specifications). YMD stores both the raw images and the corresponding image quantitative data produced by Axon’s GenePixPro. In addition, it stores descriptions of the microarray experiments. The data stored in YMD can be queried for primary data analysis. The current features of YMD include the following.

• Project management interface. This provides a four-level hierarchical data organization: projects, subprojects, experiments, and hybridizations. The user can dynamically expand and collapse a parent folder (e.g., a project, subproject, or experiment) that contains one or more child objects (e.g., subprojects, experiments, or hybridizations). Through this interface, the user can also select the datasets (hybridizations) of interest for data querying and analysis. At the hybridization level, the user can associate with it the corresponding genelist (GAL file), images (TIF files for both channels), quantitative data (GPR file), and the grid settings (GPS file). These files may be uploaded from a local computer (which will take a longer time to process) or they may be pre-uploaded in bulk to the YMD FTP server if the corresponding lab has an account set up on the server. Once uploaded, a GPR file will be unpacked into an Oracle table for efficient data querying. As a user option, the data can be normalized during the data unpacking process. Through the project management interface, the user can also select a set of hybridizations for querying and analysis. The query output can be formatted and saved locally for further local data analysis such as clustering.
• Laboratory user management and access privileges. There are five types of user roles: DBA, Lab Admin, Level 1, Level 2, Guest. The DBA user can create a lab, the PI of the lab, and guest users within the lab. The PI is automatically granted the Lab Admin role that (i) allows creation of new users within the lab (ii) assign user roles (including lab admin role), and (iii) assign project privileges to Level 2 and Guest users. Level 1 users have full access to all the projects within the lab, but such users cannot create users nor assign project privileges. Level 2 users have read or write privilege to the assigned projects. Guest users can have read-only access to the assigned projects.
• Dye Assignment. YMD allows the user to specify which florescence dye is used for the experimental versus reference sample. Typically, Cy5 is used for the experimental sample and Cy3 is used for the reference sample. However, this order may be swapped sometimes by individual users. Depending on how the user specifies the output, the ratio calculation currently performed by GenePixPro may not indicate such a dye swap. Therefore, we have incorporated into our data unpacking routine a module that takes this dye swap into account. This module calculates on-the-fly the ratios based on the dye assignment given by the user (instead of reading the ratio values directly from the GPR).
• Genelist management. This allows the user to change the properties of the previously uploaded genelists. For example, with this new interface, the user can allow new lab(s) to share the genelist(s) generated by his/her lab. Currently, YMD lets all users within a lab upload a new genelist and change the properties of existing genelists. This option may is restricted to only lab admin users).
• Linking of query results to genome annotation. Different labs that carry out microarray experiments with different organisms may want to link out to different annotation sources. An option has been created to allow the user to specify this link out at the GAL file level. The ID column in the GAL file will be used for this link out. It can be linked out to multiple sources.
• Aggregation of expression data. This allows the user to extract the expression values (corrected ratios) for all the genes across an arbitrary number of hybridization datasets. The results are outputted to a tab-delimited files with the first column representing the gene identifiers and the subsequent columns corresponding to the expression values of the individual hybridizations (experimental conditions).
• Genelist upload. YMD lets user of a given laboratory to upload a genelist (GAL file) that represents an array platform. Such a genelist file contains information about the coordinates of probe elements spotted on the array slides as well as identifiers (e.g., genbank accession numbers) of the probe elements. By default, the uploaded genelist is private to the current laboratory of the user who uploaded it. However, the user can make the genelist public (accessible by all other laboratories) or sharable by a number of designated laboratories.
• MicroArray Convolutor (MAC). This platform-independent program provides a Web interface that allows users to map coordinates between PCR plates (96-well or 384-well) and microarray slides. This interface allows the user to input the following: a) a text file that contains the plate identifiers, well locations, and DNA sample identifiers; and b) parameters that describe the arrayer’s configuration including the number of pins, dipping pattern, printing pattern, number of replicates if any, etc. Given this input parameters, the program will generate as output a tab-delimited text file corresponding a genelist.
• Keck ARray Manager and Annotator (KARMA). This Web-based program allows comparison among different array platforms (e.g., cDNA, oligonucleotides, and Affymetrix) to identify analogous DNA fragments. Comparison between platforms with the same species is based on unigene cluster IDs. When comparing platforms with different species, we use the homologene information obtained from NCBI. Currently, this program is used to compare custom slides/chips as well as micorarray slides/chips that are distributed by the Keck Microarray Resource at Yale.

Status Archives

Copyright © 2003, Yale University, New Haven, Connecticut, USA. All rights reserved.
Comments or suggestions to Greg Blasko.

Last modified: 28-Mar-2003 (GB)