Q. When will Silicon Genetics fully support the use of GeneSpring on OS X?
Since GeneSpring 5.0, Silicon Genetics now fully supports GeneSpring natively on OS X. It is strongly recommended that all OS X users upgrade to Mac OS X version 10.2.1 or 10.2.3 as these versions include the Apple 1.3.1 JVM.
Q. What measurement technologies does GeneSpring work with?
A. GeneSpring works with any genomic or proteomic technology that associates numbers with
genes. This includes microarrays, Affymetrix chips, Clontech or Research Genetics blots,
SAGE and RT-PCR. There is a comprehensive suite of normalization options appropriate for
different technologies.
Q. Is GeneSpring client/server? How do I share data with
colleagues?
A. GeneSpring is unique in that it is part of a flexible computational architecture that optimizes performance in a variety of different contexts. Researchers have the choice of using either server- or client-based computation, or a combination of the two, depending on what is best for their particular situation. The GeneSpring client can connect to Signet, a scalable repository that stores numerous forms of expression data which allows collaborative researchers to access each others data using a simple web browser. Data on Signet (which can run on multiple processors), appears as if it resides on the local computer and can be easily saved to the local hard drive. Further computational power can be added because Signet is able to farm out tasks to powerful remote servers. These flexible combinations of client-side, server-side and remote computing can be utilized to ensure maximum computational efficiency.
GeneSpring also allows data sharing through database connections. If you are using
an ODBC link to an SQL database, sharing is simple. If you are using flat files, you can use
two directories for data; a local directory, and a shared directory on a file server. You can experiment
on data in your local directory, then move it to the file server when you want to share it.
Alternatively, you can run everyone off a shared directory.
Q. Does GeneSpring work over the web?
A. GeneSpring can access data on the Web in a number of ways. It can check for annotations on a number of public databases including Genbank, Unigene, Homologene and LocusLink or connect to any web location that you specify. GeneSpring works over the web via Signet, a web-based data repository that allows you to view your analysis results from anywhere in the world. Published results can be uploaded and downloaded from the web using Signet, so that researchers in remote locations can easily collaborate on a project.
Q. What statistical analyses does GeneSpring do?
A. GeneSpring has many analyses designed for genomics:
Clustering
Gene trees (hierarchical dendrograms)
K-means (non-hierarchical)
Like a known gene or average of genes
Like a pattern drawn with the mouse
Selections of genes
Genes selected with the mouse
Genes with high confidence
Genes with relative expression in certain ranges
Comparison with existing knowledge
Automatic function / pathway assignment, model building
Comparison with results of old analyses
Comparison with results of analyses by colleagues
Pathway analysis finding genes that fit in a certain place in a pathway.
Sequence analysis to automatically find regulatory sequences.
Automatic functional annotation of sub-trees in dendrograms.
General Analysis
Find the most significant patterns in experiments (can be used for principal
component analysis)
Automatically detect skew-ness in experimental data.
Q. What are the different normalization options in GeneSpring?
A.There are 16 different normalization options in GeneSpring including intensity dependent (Lowess) normalization, global per chip or per gene normalizations, data transformations options, normalization to positive or negative controls, normalization to specified samples and more.
These may be used in any combination, although generally if you have references for
each gene you will not want to employ any of the other options. Similarly, if you are
normalizing to positive controls you generally will not want to employ the per
experiment normalization option.
Q. How much data can GeneSpring handle?
A. GeneSpring can handle as much data as you have disk space to hold. There is no limit to how many chips you can work with.
Q. How much data can GeneSpring load at once?
A. GeneSpring has a memory overhead of about 50 bytes per gene per sample. There is also additional memory required for the program, for miscellaneous data, etc. For example, on a PC with 256MB RAM GeneSpring can load over 500 samples, each measuring ten thousand genes, at the same time. On a more powerful computer, GeneSpring will load an even greater number of samples simultaneously.
Q. Under what environment does GeneSpring run best?
A. Although GeneSpring will often run with considerably less RAM, we recommend
that you install 512 MB of memory. If you are using larger genomes (such as C.
elegans or human), with many samples and many chips, you may require upwards of a gigabyte of RAM for optimum performance. In addition, Macintosh users should consider installing a two-button mouse, as many of the common functions in GeneSpring are accessed by
right-clicking/ctrl-clicking.
Q. What is a Java Virtual Machine (JVM)?
A. This is part of the operating system that manages the resources that GeneSpring uses.
Different manufacturers make different versions, with different speeds and bugs. We
recommend the Sun (J2SETM)1.4.0_02 JVM and for Mac OS 9.x, MRJ 2.2.5.
Manufacturers regularly update the Java Virtual Machines, and when they do you can install
new ones - this generally makes GeneSpring faster.
Q. What is meant by genome in GeneSpring?
A. A genome in GeneSpring represents a set of genes upon which experiments are done, and information about these genes. Generally, all the experiments and knowledge about a genome are stored in a separate directory. A GeneSpring genome need not correspond exactly to a real genome, although it usually does. For instance, a genome may contain three alleles (considered to be separate genes), whereas a particular organism would never have all three simultaneously.
Q. Which genomes does GeneSpring work with?
A. GeneSpring works with any genome. You can use either genomes set up by Silicon Genetics,
or you can easily define your own genomes. Some genomes have more functionality than
others. For instance, with a fully sequenced genome you can physically display genes
according to their exact position on the genome, and search for regulatory sequences.
Genomes that have not been sequenced but have been at least partially mapped can show
mapped physical position.
Q. Can I use a GenBank file to describe a genome, and add in some extra genes?
A. Yes, you can. This is typically done to represent a strain
slightly different from the sequenced strain. For more details on how to do this see our
Q. Can I put in my own links to web-based databases?
A. Yes. You can make a link to any web based database (like GenBank) for extra information on
genes. All that is required is that the database have a URL that contains the name of the
gene. See our for details.
Q. What is a gene list?
A. A gene list is a group of genes with some common property. This property could be
membership of a pathway, sequence motif, sub-cellular localization, interactions with certain
proteins, etc. Gene lists can come from a variety of sources - external databases, internal
databases, and calculations performed by GeneSpring. Gene lists can also have a number
associated with each gene. This can be any number. When GeneSpring makes a list for
which some numbers are appropriate, GeneSpring will automatically associate those numbers
with the experiment. For instance, when you find genes with expression patterns like a given
gene, the genes will have correlation coefficients associated with them, and will be ordered
with the most similar gene first.
