PU Computing & Deep Learning (DL) in the Research & Development of new Pharmaceutical Products

The deep learning (DL) has been linked to fields in various scientific areas, serving as a primary tool in pharmaceutical chemistry and pharmacology which are associated with drug discovery and drug development.

This application is highlighted in:

  • using modern experimental graphics processing units (GPU)
  • the development of algorithms with the capabilities of GPU.

The fields of bioinformatics, cheminformatics, and medical informatics, rely on Computer-Aided Drug Discovery (CADD), utilizing Deep Learning methods enhanced with GPUs. The pioneering applications of CADD emphasize its contribution to solving problems related to statistics and modeling. However, mechanistic learning is still under development to offer substantial benefits.

In 2007, NVIDIA introduced the Compute Unified Device Architecture (CUDA) transforming the role of the computational power of GPUs from CPU to GPU.

In September 2014, NVIDIA introduced cuDNN, a GPU – accelerated library with functions for use in deep neural networks (DNN)

Explanation of Deep Learning Models in GPU

The GPUcentered deep learning models, the core concept, is based on systematic reductions in the computational costs of models with respect to algorithms that run on CPU. At the same time, high-performance deep learning models make numerous improvements with developing the GPU.

Recently, GPU algorithms have had the ability to adapt to the needs of ecosystem operators in the energy sector. As a result, deep learning models are expected to be an important part of future biomedical phenomena, highlighting the impact of AI and quantum computing and improving parametric chromatographic methods.

The combination of the continuous and uninterrupted application of formalisms is made possible by the broad application of DL models and GPU parallel processing. Based on the hardware selection, DL is combined with the ability to solve problems by applying economic optimization and analysis based on the model of the prospect theory of risk assessment (QSAR) as the scheduled production of formal models.

Recently, there have been hybrid methods of technical knowledge where the terms combine synthetic deep learning models. The result is a large and accurate selection of chemical compounds with the help of experimental methods that produce data sets in the development of virtual libraries.

More generally, the use of Big Data enables the broad use of DL algorithms. With the automated application of the genetic algorithms and the high screening efficiency of Highthroughput screening(HTS), huge amounts of data are processed.

The use of and the application of AI models is a milestone for chromatographic and preclinical diagnostics. The automation of this diagnostic method with the help of the Technical Knowledge and the broad and with accurate experimental application with many data of DL models and GPU parallel processing gives new prospects in the continuous application of new excellent models.

CBS
Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.