A lone biology experiment can easily generate more than a gigabyte of data per day, and deep space data collection in astronomy can generate more than a terabyte of data per night, according to computational biologist Stephen H. Muggleton in a new article inspired by the "2020" report in the journal Nature. Regarding machine-generated hypotheses, he relates a working example of how his research group has "used inductive logic programming (a subdiscipline of machine learning) to discover key molecular substructures within a class of potential cancer-producing agents."

Using the same techniques, they have "been able to generate experimentally testable claims about the toxic properties of hydrazine from experimental data - in this instance, from analyses of metabolites in rat urine following low doses of the toxin."

Elementary, my dear Muggleton.

The "2020" report ends with a call to bold measures, stating, "We are in important, exciting, indeed potentially extreme, times in terms of the future of our planet, our society and our economies ... governments, scientists and policy makers cannot afford to simply wait and see what happens."

The report then makes ten recommendations: establish science and science-based innovation at the top of the political agenda; urgently rethink how we educate tomorrow's scientists; engage the public in science; rethink science funding and science policy structures; create new kinds of research institutes; reenergize computer science to tackle "grand challenges"; develop new conceptual and technological tools; develop innovative public/private partnerships to accelerate science-based innovation; find better mechanisms to create value from intellectual property; and last but not least, the group encourages one and all to "use our findings."

That last bit is very human.

Microsoft Research Cambridge will further encourage the report's agenda by making 2.5 million euro available to support new research that specifically addresses the areas outlined within "Towards 2020 Science."