The spring meeting of the ACS saw Cresset presenting a comparison of Spark, with a quantum mechanical approach NEAT that was published last year, and discussion on the best way to prepare datasets protesting virtual screening and the development of the BlazeGPU.
Tim Cheeseright presented a detailed study on the use of Spark for scaffold hopping on Sildenafil. He showed that using the XED force field gave comparable results to those published by Pfizer. He speculated that this was a combination of working in product space and the unique abilities of Cresset’s force field to model aromatic species. The full presentation is accessible below.
Tim also presented work that we have been doing on the assessment of virtual screening performance. Mark Mackey is now writing this up as a paper and hence the slides are not currently available. If you would like to discuss our thoughts then please get in touch.
Simon Krige presented the work he has been doing on the conversion of Cresset’s core alignment technology from Fortran to openCL. He outlined the many advantages of working in a language that enables code to run on both central and graphics processing units. He showed Cresset’s virtual screening system, Blaze, has been speeded up by a factor of 45 when running on the latest generation of accelerated device for Nvidia. The full presentation is shown below.
We will be discsusing aspects of these presentations at our european user meeting in Cambridge, June 20,21, please join us!