Forum:How can I illustrate the importance of open source software in bioinformatics for people outside the field?

Question

[Keith Callenberg]

I think it's safe to say that the majority of people in science, technology, and engineering fields are now comfortable with the idea of open source software. But this wasn't always the case, and there are still many people out there that are not comfortable with the idea of open source software being relied on for important, mission-critical tasks. Even if an open source package has been shown to be dominant in terms of accuracy or performance or some other metric, there are those that would choose a paid licensed, closed-source alternative simply because it has a company name and a support contract standing by.

Someone entering this field, starting to pick up new skills, would discover soon enough that it is hard, if not impossible, to get by in bioinformatics without open source tools. But say you are communicating with a consultant, or someone in a IT, technical support, legal, or otherwise auxiliary role. Are there any quick statistics or objective explanations to help convince them that they should not waste time avoiding open source?

Specifically a number that might help sway skeptics is the percentage of biomedical organizations that utilize open source tools somewhere in their analysis. Is there any way to even roughly estimate that value, or something similar?

I've found that subjective arguments about open source's ability to track bugs and collaborate with a variety of developers with different use cases have not worked, so it would be great to have some objective facts that make the case clear.

[Keith Callenberg]

I have trouble even finding paid, closed-source packages for genomics. (That alone could be an argument.) But for example, I have seen people outside the field place higher value on packages like CLC Bio because the company "owns" and supports the tools contained within.

Answer 1

[harold.smith.tarheel]

You can direct skeptics to the SEQanswers software wiki, which includes citation metrics for various software packages (open source and commercial). No commercial package cracks the top 25 (it looks like LaserGene is the first at #28), and citation numbers are orders-of-magnitude lower.

Also, many commercial packages license open-source tools for the analysis. For example, Genomatix Genome Analyzer uses MACS for ChIP-Seq peak calling.

Answer 2

[William]

Emphasize that you are using cost effective proven bio-bioinformatics tools that hold the majority market share and that guarantee that you have full acces / ownership of the produced data (trough open file formats & API).

Most commercial tools come together with a vendor lock-in through the use of proprietary file formats / API, to force a yearly license cost for retaining the right to acces/use the data. You will have produced a lot of hopefully valuable data in a few years and you don't now what the price wil be in a few years to buy the right from the software provider to acces / use that data.

And that vendor lock-in limits you in the ability to build a larger system by not being able to pick and chose the best bio-bioinformatics tools and combining them. If you want more functionality you have to go to the same commercial provider and ask them if you can pay them to provide the new functionality. And they might not know anything about this functionality.

This proprietary file formats / API point can be very convincing for choosing open source software or at least make you less locked-in to the commercial software by having them support open file formats and API's (this might increase the yearly license cost because they don't have a guaranteed multi year + increasing income trough the lock-in).

The open file formats / API also makes switching to another commercial or open source tool much easier and makes it easier to build larger systems where you need specialty tools / domain knowledge from different software providers.

Answer 3

[LLTommy]

Even though it might not speak directly about 'open source' software, it is intersting read and might give you some arguments for why openess of software and data is beneficial.

http://www.ebi.ac.uk/about/news/press-releases/value-and-impact-of-the-european-bioinformatics-institute