SLAS

MP31:KevinWilliams:TheRobotScientistEve

From LabAutopedia

Jump to: navigation, search

Kevin Williams1, Elizabeth Bilsland2, Stephen G Oliver2, Ross D King1

1. Department of Computer Science, Aberystwyth University, Aberystwyth, Ceredigion, United Kingdom
2. Department of Biochemistry and Cambridge Systems Biology Centre, University of Cambridge, Cambridge, United Kingdom

Robot Scientist Eve: Selection and confirmation of chemical compounds with potential for activity against parasites causing neglected tropical diseases


Abstract

Robot Scientist Eve has been developed to screen for drugs specifically targeting parasites responsible for neglected tropical diseases [1].

Genetically engineered yeast (S. cerevisiae) strains were developed for the project; these have yeast genes replaced with either homologous target genes from a parasite (parasite mimetic) or from the human host (human mimetic) [2]. Each yeast strain is labelled with specific fluorophores to allow measurement of growth by UV/Vis fluorescence. The assay consists of comparative growth of human and parasite mimetic yeast strains. Amongst the diseases being studied are malaria (wild type and drug resistant forms of Plasmodium falciparum and P.vivax), sleeping sickness (Trypanosoma brucei), Leishmaniasis (Leishmania major), Schistosomiasis (Schistosoma mansoni) and Chagas disease (Trypanosoma cruzi).

An experiment typically consists of screening three modified yeast strains growing in competition in 384 well plates in the presence of the test compounds. Hit compounds are identified by suppression of growth of the parasite mimetic yeast strain relative to that of the human mimetic version. Machine Learning rules for Eve have been derived on this basis.

Growth curves of each strain are acquired for each well; the Robot Scientist Eve then identifies hit compounds from these data automatically, eliminating toxic, auto-fluorescent or fluorophore specific compounds. Eve then executes and analyses a second screen (validation screen) at multiple drug concentrations.

The work has provided robust and reproducible means of analyzing the in vivo activity and specificity of very diverse compounds. Several promising compounds have been identified, and these are now being validated by alternative techniques such as in vitro binding assays and using intact parasites.

The screening process is being further developed to allow Eve to learn from its results during the course of an experiment; this will provide a basis for prediction of the behaviour of untested chemicals, and lead to the selection of promising drugs from a larger universe of compounds.

[1] R. D. King et al, The Automation of Science, Science 324 (2009), no 5923, 85-89
[2] Bilsland E, Pir P, Gutteridge A, Johns A, King RD, Oliver SG (2011) Functional expression of parasite drug targets and their human orthologs in yeast. PLoS Neglected Tropical Diseases e1320 (9 pages)

Click [+] for other articles on 
SLAS2012 Posters(1 C)
The Market Place for Lab Automation & Screening  The Market Place