Selecting Information
From The Endless Results
|
|
SIFTER is intended to assist in managing the massive amounts of data that are generated by groups doing genetic analysis. In the project where SIFTER was conceived, it was common to generate over 12,000 analysis results for their bi-annual meetings. With this amount of data, even the simplest activities require great effort.
Thus SIFTER was conceived to deal with problems in handling data on this scale. The results are saved in a database and may be searched looking for "interesting" results which can be conveniently displayed. The convenience of this approach facilitates remote collaboration and allows the analysts to see the results in context. At the same time SIFTER can be used to integrate a group's results and maps with other resources found on the Net.
SIFTER is a repository for genetic maps and analysis results. Tools are provided to scan common analysis results or maps and insert the data into the database. One may also write extensions to the software to insert data in formats that are not supported by SIFTER.
Once the data is loaded in the database, one may make queries to select results or maps that are of interest and then display them individually or as overlays on each other. SIFTER data is organized into separate projects which correspond to organizational groups doing analysis. Each project should have it's own database as well as its own area in the web server (i.e. URL).
Conceptually this is a simple and common process, however, the scale of these data presents complications. Doing a query which results in a list of hundreds of results is pretty much useless as the interesting results are lost in the uninteresting.
To help with this process, SIFTER defines a set of attributes to help describe the results so it may be more easily searched. In addition the local analyst may define his own attributes and assign them values of his choosing.
For instance, SIFTER defines attributes like chromosome, analysis type (e.g. association, linkage etc.), map date, analyst name and others. The local analyst could invent other attributes that help organize the results according in other innovative ways, like 'event' (meaning results done for a particular meeting).
SIFTER is a Unix application and consists of several major components:
Depending on the group, each of these components could involve several individuals with very different skills. The last of these is the end user for SIFTER, namely the analyst interested in searching the results. The first three will be referred to as administrators, namely:
database administrator - one who manages a relational database. SIFTER requires simple database and tables. In many environments the database will be provided by some database administrator for the organization.
web administrator - generally a web master or one who understands and has the authority to install web pages and CGI scripts. We assume there is an already functioning web server. The web administrator must set up the web server:
data administrator - one who loads the maps and results into the database. In some instances, this might be the analyst while in other larger projects, this might be a separate individual. This requires a thourough understanding of the formats of the various files and attributes to be used for the project.
Complete details on the installation and use of SIFTER may be found in the following:
This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License (http://www.gnu.org/copyleft/gpl.html) for more details.
This software was originally developed by
The developers are grateful to a number of people who have contributed their support and input: