Support Vector Machine – a model for QSAR still robust in the era of Deep Learning
The new QSAR framework in Flare makes generating and using 3D-QSAR descriptors to build models with good predictivity and generalizability ...
Welwyn Garden City, UK – 12 June 2013 – Cresset, innovative provider of chemistry software and services, announces the release of the GPU version of Blaze, the leading virtual screening platform. BlazeGPU runs at 40 times the speed of Blaze using affordable consumer hardware, making this effective lead generation tool accessible to any research lab.
This dramatic speed increase means that screening a database of a few million compounds can be achieved overnight using a single desktop box with 4 GPUs, rather than requiring an expensive Linux cluster.
Alternatively, a small cluster equipped with GPU coprocessors can now screen virtual libraries of tens or hundreds of millions of molecules using 3D shape and electrostatic comparisons. This database size was previously accessible only to 2D methods.
“We have achieved a fantastic 40x speed-up of Blaze while losing nothing in scientific accuracy or thoroughness,” says Dr Robert Scoffin, CEO of Cresset. “BlazeGPU opens the way to new experiments that would previously have been too computationally costly to run, such as assessing the 3D diversity of a collection of thousands of compounds.”
“The release of BlazeGPU is the culmination of Cresset’s 18 month project to port our core algorithms to these exciting devices,” says Dr Mark Mackey, CSO . “Critical to the implementation is the fact that BlazeGPU