A robot scientist, which researchers at the Universities of Manchester, Cambridge and Aberystwyth call ‘Eve,’ could make drug discovery much more efficient and cost-effective. Eve proved her potential by discovering on her own that an anti-cancer compound could also be used in combating malaria. In the journal Interface, the researchers state that the use of robot-scientists is an extension of the growing trend towards greater automation in science. Automated scientists are able to develop and validate hypotheses. They can also use laboratory robotics to perform experiments, analyze their results and amend their hypotheses. Image: Interface. Eve is also extremely adept at recording scientific knowledge. Because the experiment are automatically conceived and run by computers, it can capture every step of the scientific process and digitally curate them. Adam was the first robotic scientist, followed by Eve. Adam was created at Harvard and Aberystwyth. Eve was created by the same team who developed Adam. Eve is designed to speed up drug discovery and make it more affordable. Researchers describe the ways Eve could help to identify new drugs for neglected tropical diseases like Chagas’s disease or African sleeping disorder. Ross King is a Professor at Manchester Institute of Biotechnology. He said that neglected tropical diseases have become a major problem in the world. They kill millions of people each year and are infected with hundreds of million of people. These diseases are well-understood and we have the ability to attack them with small-molecule drugs. They are not attractive to the pharmaceutical sector due to their cost, speed and economic return. Eve uses its artificial intelligence to draw from her early success stories and choose compounds with a high chance of acting against the selected drug target. It uses a smart screening method that is based on genetically engineered yeast. “This allows Eve not to select compounds toxic to cells, but to choose those that will block the action. Any equivalent human protein remains intact.” It reduces costs, uncertainty and time required for drug screening and can potentially improve millions of lives worldwide. Ross Donald King, a professor of machine intelligence at the School of Computer Science at Manchester, works at the Manchester Institute of Biotechnology and Machine Learning and Optimization group. Image: Wikipedia. The robot scientist was designed to assist in early stage drug design. Eve begins by testing every member of a large group of compounds using the brute force method of mass screening. These compounds are tested against the ability to perform automated engineering tests (tests and assays). They can be performed much faster and cheaper than custom-made assays. This allows for more assays to be used and screening equipment can be utilized more effectively, increasing the profitability of any discovery made within a budget. The robot system Eve can screen many 10,000 compounds every day. Mass screening can be slow, tedious and time-consuming because every compound in the library has been tested. In the sense it does not make use of information learned during screening, this isn’t intelligent. Eve randomly selects from the library a portion to expedite the screening process. To reduce false positives, promising hits are repeated several times. Eve takes the confirmed hits, and then uses machine learning and statistics to find new structures. Although she is currently not able to create such compounds, the researchers believe this could change in the future. Steve Oliver works in the Cambridge Systems Biology Centre as well as the Department of Biochemistry. He said that automation has benefited every industry, and science is no different. Machine learning could be used to speed up science and reap enormous rewards. The scientists created assays that target key compounds from parasites, such as Chagas disease, Schistosomiasis, and malaria. They tested these against a collection of approximately 1 approved clinical compounds. Eve was able to demonstrate that a previously investigated anti-cancer drug inhibited a malaria parasite-specific molecule, DHFR. The robot scientist discovered a new use for an already existing compound. Medications which inhibit DHFR can be used to prevent malaria and they are given to over one million children. New resistant strains of parasites are now being discovered, which makes the need for more drugs urgent. Professor Oliver said that no new antimalarial has yet been found to target DHFR, and it was not possible to test this in clinical trials. Eve’s breakthrough could have far-reaching implications beyond proving a new method of drug discovery. The European Commission funded the research. Citation: “Cheaper and faster drug development, validated through the repositioning drugs against neglected tropical disease,” Kevin Williams. Elizabeth Bilsland. Andrew Sparkes. Kurt De Grave. Jan Ramon. Michaela de Clare. Worachart Sirawaraporn. Stephen G. Oliver. Ross D. King. Interface. Published February 2015. DOI: 10.1098/rsif.2014. 1289 Artificial Intelligence Robots are equipped with what computer scientists and roboticists call “AI”. The software makes them think and behave just like humans.
