Support for discovery and verification of candidate materials
ėŽė§ íë LG CNS has completed the construction of Dong-A Socio Group’s “AI New Drug Development Platform.” This platform, which was established with Dong-A Socio Group’s IT affiliate DAI, integrates and standardizes distributed new drug research data and supports major research processes from candidate discovery to verification with AI.
LG CNS announced on the 12th that it has completed the construction of Dong-A Socio Group’s “Artificial Intelligence (AI) New Drug Development Platform” with Dong-A Socio Group IT affiliate DAI.
This platform, which took about six months to build, integrates and standardizes scattered new drug research data and supports the entire process of new drug development with AI from candidate material discovery to verification. In particular, it has a structure that continuously compares and learns AI prediction results with actual experimental data to increase prediction accuracy as research progresses.
Usually, it takes 10 to 15 years and a huge amount of money to develop a new drug. This is because numerous candidate materials must be verified one by one. LG CNS explained that this platform can reduce the development period, cost, and risk of failure by quickly selecting candidate materials that are highly likely to succeed and predicting their effectiveness and safety in advance.
The platform gathered scattered data such as compound and dielectric information, experimental results, papers, and patents in one place. Researchers can immediately utilize AI analysis and prediction functions by fetching necessary data within a single platform without traveling to and from a separate system.
There is also a support method for each stage of new drug development. In the stage of identifying the cause of the disease, genetic information for each cell and location information within the tissue are analyzed and visualized with AI to help discover therapeutic targets. In the candidate material design stage, the Generative AI designs a new material structure that meets the conditions, and in the verification stage, the possibility of binding between the candidate material and the target and the stability of action are predicted by simulation to select promising candidates. It is a method of virtual verification first in a computer environment prior to an actual experiment.
By comparing AI prediction results with actual experimental data, we have also prepared a structure to retrain the model so that the prediction performance will continue to improve as the data accumulates. In addition, considering the characteristics of sensitive new drug research data, a management system was also established to respond to pharmaceutical industry regulations and security requirements.
Meanwhile, LG CNS announced that this achievement was based on the pharmaceutical and bio AX capabilities that have been accumulated so far. Previously, he participated in the Ministry of Health and Welfare’s “K-AI New Drug Development Preclinical and Clinical Model Development Project (R&D)” and has accumulated results in pharmaceutical and bio AX projects in the public and private sectors by developing Chong Kun Dang’s “agentic AI-based annual quality evaluation report preparation service.” Recently, it is expanding its business through ‘Agentics Works for Bio’, an agentic AI platform specializing in pharmaceuticals and bio.
“We will contribute to innovation in research and development of new AI drugs in the domestic pharmaceutical and bio industries through differentiated AX technologies including AGENTIC AI and specialized capabilities in the pharmaceutical and bio sectors,” said Lee Jae-seung, head of LG CNS’ cloud business. “We will actively support customers to strengthen their competitiveness in developing new drugs.”