Overview

The GCG Wisconsin Package is an integrated package featuring a comprehensive collection of DNA-, RNA-, and protein-sequence-analysis tools. Its functions span the range of sequence-analysis tasks from sequencing-run assembly, sequence editing, database searching, and pattern finding to nucleotide-chain folding. The Wisconsin Package is to the sequence-analysis world as Photoshop is to the graphics-editing world, Mathematica to the symbolic-mathematics world, and PowerPoint to the desktop-presentation world - a de facto standard that most serious users expect to be present, and some can scarcely function without.

The Wisconsin Package's greatest strength is that it is a collection of relevant pieces of software gathered into a suite. In-house programs, as well as software and algorithms adopted from other researchers, are made available under a unified interface. When new and useful software is published, Genetics Computer Group licenses, adapts, and adopts the software into the Wisconsin Package. Recent examples include the MEME/MAST suite of de novo pattern-discovery programs, the Staden sequence-analysis package, the HMMER Markov-model construction suite, and the PFAM protein-motif database. This umbrella approach to providing sequence-analysis tools removes the need for the constant collection of new software by the end user, and the requirement for users to learn new interfaces as new software is released. A frequently overlooked aspect is that the components of the Wisconsin Package are all functional as command-line programs in the Unix environment, which allows the user to automate complex tasks. Unfortunately, this feature is often considered to be an impediment by beginning users, who prefer to use the X Window-based SeqLab interface. For new users, investing as much effort in learning the Unix environment as in learning the Wisconsin Package will pay off in considerably more than doubled efficiency in their use of the software.

How Long Did It Take to Learn to Use It Productively?

The Wisconsin Package takes a day to learn, and considerably more time to master. Most novice users usually can learn to use any given component of the package to solve a particular problem in an hour or so. Learning all the optional parameters that can be used to control a component takes longer. With devoted study, it's possible to make rapid progress at learning the system as a whole, but the practical reality is that many users take months before they begin to learn to link the many components together and to automate repetitive tasks. An additional layer of complexity in the learning curve is the requirement to have a working knowledge of Unix to make full use of the system.

Product Quality

Customizability

Highly customizable, partly through customization of the program, and partly through the inherent customizability of Unix itself.

Ability to Program in Scripts, Add Extension Modules, etc.

As a collection of independent pieces of software under Unix, the Wisconsin Package suite of programs submits very nicely to control through shell scripting, the insertion of custom modules, and the addition of user-written routines.

Ability to Import and Export in Different File Formats

The majority of the programs in the Wisconsin Package use an internal, GCG-specific set of file formats, but there is a "reformat" command that reads a plethora of sequence formats and rewrites the data into GCG format. Output options are somewhat more limited. Users should be aware that GCG-supplied versions of various adopted software packages such as FASTA, have been adapted to read only the GCG format, and not the file formats that they were originally designed to use.

Useful or Unusual Features

The database subscription, an optional add-on to the basic software package, allows fast local searches of all the major public databases and is one of the most noticeable timesaving features. The real power of the GCG Wisconsin Package, however, lies in a feature that many users seem to try to avoid - the fact that it is a cooperating collection of independent pieces of software operating on top of the Unix operating system. This allows the operation of the software to be customized and extended in a way that cannot be matched by any desktop-based system.

As an example of the flexibility and power that this approach to software provides, consider a researcher examining a set of genes that appear by expression analysis to be coordinately regulated. If the researcher is interested in examining the regulation scheme and whether it is conserved outside the organism in which it was observed, GCG commands can be combined to help answer the question. The MEME motif-finding program can be used to search the upstream regions of the genes for potential regulatory motifs. The results of this search can then be fed into the MotifSearch database-search tool to find other genes in the nonredundant database that share these potential regulatory motifs. The genes found to share the motifs can be retrieved from the database and aligned with the PileUp command, pairwise distances extracted from the alignment with the Distances command, and the final results fed into the GrowTree command to see whether the genes appear to fall into related groups, possibly indicating conservation of regulation in a metabolic pathway.

At first glance, that example may seem a somewhat daunting collection of tasks, and the reader may imagine a considerable amount of pointing, clicking, typing, and general unhappiness to be associated with the process. Fortunately, although it seems daunting, the reality of the matter is that through the power of Unix and the integrated design of the programs provided by the Wisconsin Package, this process can actually be condensed to a series of commands on a single Unix command line. For the user who needs to do similar analyses repeatedly, application of only the slightest bit of Unix shell programming would condense it all down to a single reusable command.

Limitations

As might be expected, since the Wisconsin Package is primarily a command-line-based suite of programs, the GUI interface does not give the user access to the complete power of the command line, nor to the complete set of programs. A limitation that will not be expected, and one that is a continuing source of annoyance to many users, is that many programs in the suite place a limit on the maximum size of the sequences that they can handle (350 Kb) and require that longer sequences be broken up into segments under this length. GCG reports tha this limitation will be removed in version 11.

Comparisons with Similar Software

The GCG Wisconsin Package really has little competition that could be considered similar. Packages such as the DNAStar software suite available for Mac OS and Windows, while they perform a number of similar tasks and are designed around a more integrated interface, do not provide the flexibility or automation that make the Wisconsin Package so powerful. Other Unix-based sequence-analysis packages are not as complete.

Technical Support and Documentation

The documentation is styled in a form similar to Unix man pages, being mostly reference, with minimal tutorial or example information. Online searchable access to the documentation is included, although oddly it is not available through the conventional Unix "man" command. Genetics Computer Group personnel have always been very helpful, and they should be willing to assist in choosing appropriate hardware to meet the intended level of use. Technical support seems willing to assist with the creation of custom scripts to help automate individual end-user projects.

Target Users

Although the Wisconsin Package works well in a teaching environment, the typical user is a researcher needing powerful and highly configurable sequence-analysis software. The expense and learning curve do not lend themselves to one-time or casual use, being more appropriate for users looking for a long-term commitment to a powerful package.