A rant about data quality: machines first, humans second…

GraphicalAbstractRecently one of my papers emerged through the publication system of Journal of Cheminformatics, entitled “Machines first, humans second: on the importance of algorithmic interpretation of open chemistry data“, co-authored with Antony Williams and Sean Ekins, and incorporated into the JC Bradley Memorial Issue. Spoiler alert: the paper is about how if you’re publishing open lab notebook data without adhering to rigorously defined standards for machine readability, then you’re mostly wasting your time, and arguably making the open data situation even worse than it already is. The tone of the article is a bit less polite than I normally try to be, so fair warning, but it’s all for a good cause.

Mobile Bayesian modelling: importing

importbayesOne more key piece is in place for Bayesian modelling with apps: MMDS 1.6.1 just got approved on the AppStore, and brings with it the ability to recognise files with the .bayesian extension (or MIME type chemical/x-bayesian), and import them into the collection of available models that can be used to calculate properties. At the present time the only official way to create such models is to use the bleeding edge build of the Chemical Development Kit and roll your own wrapper code, but we’re working on that!