Research code can be messy stuff and tweaked on the fly to examine different hypotheses, writes Iddo Friedberg at Byte Size Biology. "Research coding is not done with the purpose of being robust, or reusable, or long-lived in development and versioning repositories," he says. "Upgrades are not provided and the product, such as it is, is definitely not user-friendly for public consumption." However, as a Twitter conversation Friedberg had points out, the code could have bugs or errors that affect results. Friedberg notes that the Bioinformatic Testing Consortium, of which he is part, aims to test such code. "A great idea, and if taken up by journals, having your BTC-approved code accompanying your paper would go a long way to validating the science presented in your research article," he says.
Aug 28, 2012