Contents

General Issues

Copyright and Disclaimer

You are welcome to print copies of this page for private use, and to link to it from your own Web pages to this page. But please do not make any electronic copies and publish them on your Web page or elsewhere.

For the litigious among us, all the the information on this web site is offered without warranty of any kind. The comments expressed on this site are intended as outlines of BioinfoTools' strategy and policies, not as legally binding statements. While BioinfoTools will endeavour to met these, BioinfoTools cannot be held to any of the statements made on this web site. If you wish to offer corrections or comments, these may be emailed to webmaster@bioinfotools.com

[jump to table of contents]

Introduction

BioinfoTools develops advanced bioinformatics software development tools and methods for bioinformatics. Contract development and training requests will also be considered.

BioinfoTools' products have two main aims. The first is to remove the "how" problem in developing bioinformatics software, so that customers' bioinformatics development teams can concentrate more on what bioinformatics they want to do rather than on exactly how to implement it. This aim is tackled by developing bioinformatics computing "engines" for use by customers.

The second main aim is to present advanced bioinformatics methods to better exploit customers' hard-won data. These end-user tools also serve as rigorous "real-world" tests of BioinfoTools' development tools.

BioinfoTools aims to assist customers to develop their own bioinformatics strategy, not offer customers software tools with a pre-defined set of tasks they can tackle. Answers elsewhere in this FAQ explore this in more detail.

BioinfoTools takes a "no bullshit" approach to issues with its products and support. This policy reflects Dr. Jacobs' experience with software firms which are less than forthcoming about issues with their products compared to those which openly state any known issues and offer temporary work-arounds. BioinfoTools' aim is to assist bioinformatics software development: withholding information about products hinders that aim.

In the FAQ answers below, 'end-user' means an individual using a product for biological data processing. This is distinct from a developer using BioinfoTools' products to construct their own local software. Likewise a 'customer' refers to a person or institution which has purchased a product from BioinfoTools. Customer is used to distinguish from the term 'client' which is only used in the context of the distributed computing term 'client/server' to avoid ambiguity.

[jump to table of contents]

Who are BioinfoTools' intended clientele?

Companies and academic research groups with bioinformatics development teams of any size. Bioinfotools' development products are aimed at bioinformatics staff with some software development experience (however limited), not "naive" end-users with no bioinformatics or computing experience.

BioinfoTools' end-user tools targetting specific bioinformatics problems are intended to be used by bioinformatics researchers without software development experience.

Substantial documentation will be available for all products. Developer-level documentation will include examples of use of the development tools.

BioinfoTools is primarily aimed at assisting developers to quickly develop their own bioinformatics tools/suites. It is the customer company's responsibility to provide end-user support for any products they develop using BioinfoTools' products.

[jump to table of contents]

Is end-user support available?

To repeat from the previous FAQ: BioinfoTools is primarily aimed at assisting developers to quickly develop their own bioinformatics tools/suites. It is the customer company's responsibility to provide end-user support for any products they develop using BioinfoTools' products.

Bioinfotools provides support for developers in the form of substantial documentation and demonstratory code. If the documentation clearly does not explain an issue, developers are welcome to email BioinfoTools, but they are expected to first attempt to resolve the issue from the documentation supplied. This is the traditional situation for development tools. While sympathising with those "going up the learning curve", in general BioinfoTools cannot address basic computing and computer programming skills — some basic compentence is assumed.

End-user support is limited to those products which explicitly state that this is available. This reflects BioinfoTools' main intended clientele which are the bioinformatics development teams, not the users these teams serve. (It is possible that this policy will change with time.)

Feedback on BioinfoTools' products is most welcome. Please realise time may limit responses, but you are assured that all email will be read and your points will be considered.

[jump to table of contents]

Will BioinfoTools develop software on request?

Bioinfotools is open such requests. Requests may cover any area of bioinformatics, but, in general, request falling into the areas of expertise covered elsewhere in this web site will be looked on more favourably. Briefly, these include structural genomics, protein-sequence structure matching, protein structure prediction, analysing protein sequence/structure families, protein-DNA and protein-RNA interaction predictions.

Topics outside these areas will be considered, but in most cases customers will be required to pay additional fees to cover the extra research time involved.

[jump to table of contents]

What experience do BionfoTools staff have?

BioinfoTools' is built around Dr. Grant Jacobs, an independent computational biology software developer with over ten years hands-on experience in bioinformatics software development using a range of computer languages and operating systems. His training includes an undergraduate degree in both computer science and biology (New Zealand), a doctorate in computational structural biology (MRC Laboratory of Molecular Biology, Cambridge, England) and over five years research experience as a bioinformatics scientist.

[jump to table of contents]

Do we need to cite our use of BioinfoTools products if publishing work using them?

Yes. A list of publications to cite will be made available on-line. The publications to cite may not make the initial release of any given product, but but will be included in later releases. In this case, the publication list will be forwarded to customers after purchase.

[jump to table of contents]

Products

What kind of bioinformatics software does BioinfoTools develop?

BioinfoTools is developing two types of software: bioinformatics software development tools and bioinformatics end-user tools. The latter provide cutting-edge bioinformatics solutions to specific bioinformatics problems and are discussed in the FAQ What areas of biology do BioinfoTools' software target?. Here we discuss the development tools. As said in the executive overview, this involves a '"how" to what"' shift in bioinformatics software development.

BioinfoTools is developing computing "engines" which store and process biological sequence and atomic structural data. These allow programmers to load and manipulate the data supported by that engine (eg. sequence data, alignments, atomic structural data) without having to know the detailed computer methods used to make these manipulations possible. This allows bioinformatics development staff to focus on what bioinformatics processes they wish to happen, rather than how to make this happen. More details will be available nearer to the release of these development tools. You may wish to register to be emailed when more details are available using the customer inquiry form.

[jump to table of contents]

What areas of biology do BioinfoTools' software target?

The answer needs to separately address development and end-user products.

The development tools are designed to be "generic" — applicable to a wide range of tasks on a given type of data. They are designed to assist customers to process their data as they see fit, rather than be constrained to a particular set of tasks, as is more typically the case.

Broadly speaking, BioinfoTools user-end tools are focused on "first principles" biology and as such use structural biology as the origin of much of the methods. These tools have application in structural genomics and functional biology, especially when applied to particular families of proteins of interest. More specifically, BioinfoTools focuses its interests to a modest range of areas in order to maintain its aim of providing world-class software, in particular:

     • prediction of substrate-binding sites
     • protein sequence-structure matching and structure prediction
     • protein-substrate interactions (including protein-DNA and protein-RNA)

Areas opened up from developments in the above will also also considered; after all, bioinformatics is a rapidly evolving field.

[jump to table of contents]

Why does BioinfoTools limit itself to such a narrow portion of computational biology?

It enables BioinfoTools to develop world-class software in these areas. Many companies provide bioinformatics tools which are either

     • collections of methods developed by other scientists (generally academics who have not protected their work) which they do not have in-depth knowledge of, or
     • efforts aimed at "picking the easy fruit", ie. methods which provide some basic functionality for a given problem with the emphasis on a pretty user interface or "ease of use" for end-users.

BioinfoTools restricts itself so that it can cover the literature in its specialist areas in depth. BioinfoTools is aimed at providing expert-level functionality, not "quick fixes".

[jump to table of contents]

What platforms are BioinfoTools products available for?

BioinfoTools' products are intended to be served from a wide range of Unix and Unix-related operating systems (eg. Linux, Mac OS X) and accessed from any client platform (Mac OS, Windows, Unix-based). Initial development is aimed at meeting smaller customer's needs (universities, smaller companies). Hence initial development is on Linux and Mac OS X-based systems. From there they are ported to other Unixes based on larger systems.

This choice of primary target platforms also reflects a conviction that computing is moving towards making better use of networked desktop computers, which are otherwise unused much of the time. Small companies and universities stand to gain the most from this use of networked computers.

[jump to table of contents]

Product quality

Will customers be informed of any weaknesses in a given BioinfoTools product?

Yes. If a product has a known weakness or is unsuitable for some uses, customers will be informed in the literature about that product. This information will be both on this web site and in the documentation which is part of the product itself. BionfoTools strongly believes this openness is important. Without it, sensible development of "higher level functions" built upon our products cannot take place. Dr. Jacobs' personal experience has shown that companies who "hide" weaknesses in their software only waste developer's time. BioinfoTools' aim is to aid software development, not hamper it. Thus, the more customer know about BioinfoTools' products, the better.

[jump to table of contents]

How are BioinfoTools products tested?

BioinfoTools' products are extensively tested in-house prior to release. There are four main forms of testing: (1) testing during development, (2) regression/installation tests, (3) testing via 'real world' application and (4) beta testing. This high level of testing and verification is one of the prime differences between academic software development projects and quality commercial software development.

As the products are developed, testing code is produced in parallel. Each new addition to the product has corresponding test(s) added to the test code. After the addition of each new feature, the complete testing code is re-run.

A regression test will be supplied with each product which will test that your local installation performs as desired. Regression tests are large test suites designed to exercise all coded procedures (methods, API calls) within a product and as many combinations of these procedures as possible, given a realistic amount of time. These regression tests are developed from the testing code created during product development, so that the regression test for a product evolves as the product is developed.

In another form of testing, all BioinfoTools software development products are used to solve "real world" bioinformatics problems before they are offerred to customers. This means BioinfoTools products are tested in the manner that customers would use them.

Finally, some products may undergo beta testing. In beta testing, the complete product is released in advance to a limited number of customers under a non-disclosure agreement in exchange for feedback on any bugs, etc. If you register with BioinfoTools, you will be informed of any opportunities to act as a beta tester. In the first instance beta testers will be chosen from biotechnology companies and research institutes within New Zealand.

[jump to table of contents]

Will BioinfoTools guarantee the performance its software?

While the core engines and primary APIs (in and off themselves) developed by BioinfoTools are unlikely to have major bugs due to the way they are developed, like all software development companies, we cannot guarantee they will work.

Software development is simply too complex to apply a rigid guarantee of its functionality. The reliability of software depends on much more than the code written by programming staff. It also depends on the software development tools used by the programmers (eg. compilers, debuggers) and on the computer hardware and computer operating system the products are run on. These products are developed by other vendors and we cannot be responsible for errors in their products.

BioinfoTools does endeavour to prove the correctness its core algorithms, both using theorhetical techniques (eg. logical algebra) and practical tools (eg. our in-house and distributed regression tests). It is strongly advised that customers regression-test their installed software; this test will provide a strong test that the software works on your platform.

[jump to table of contents]

Licensing Issues

Licensing details will be presented as products are released. There will be academic licenses in addition to commercial licenses. For most products a range of licenses will be available to match the customer, from single user academic licenses to large site commercial licenses. In addition, it is intended that there be embedded distribution licenses to allow customers to distribute software with BioinfoTools software "embedded" within it.

[jump to table of contents]