Viva Biotech's computational chemistry is based on physical chemistry models and empowered by artificial intelligence algorithms. With the aid of supercomputing clusters, it has been widely used in various drug R&D stages. How has Viva's CADD platform been implemented to drive drug development? How can the current challenges in PROTAC molecule design be addressed? How does the team collaborate efficiently with medicinal chemistry to drive drug design and optimize the process? Recently, Viva interviewed Dr. Qian Yue, Senior Director of the Biology Department, who is a senior expert in the field of CADD. She shared her rich experience and insights with us.
Q1：Recently, ChatGPT’s popularity has swept the world and powerful computing tools have become indispensable. With the AI boom, Dr. Qian, what impact has CADD, which has existed for more than 30 years, had in various aspects of new drug research and development? What impact will the emergence of AIDD have on CADD?
Dr. Qian: Computer-aided drug design (CADD) has transformed the drug discovery process from purely serendipitous to a more rational process. It serves as the catalyst of the more efficient and accurate development of novel drugs in the past decades. Although in silico methods have been extensively used in various drug discovery stages, AIDD tools can be particularly useful in fields where there is insufficient knowledge about the underlying biological mechanisms. Such models would benefit greatly from the rapid accumulation of high-quality experimental measurements and offer valuable insights into the drug discovery process in conjunction with the conventional physics-based approaches.
Q2：Protein degradation technology has developed rapidly in recent years due to its ability to induce the degradation of pathogenic target proteins. However, current PROTAC technology still faces the limitations of difficult-to-drug targets and E3 ligases that can be applied to molecular design. Regarding this, how does Viva’s CADD address the design challenges of PROTAC molecules?
Dr. Qian: We implement cutting-edge techniques in generating the ternary complex structures. This includes PROTAC conformation sampling, protein-protein docking as well as the structure mapping algorithm. A series of filters and scoring functions are applied during the process to scrutinize the candidate structures which lead to a high hit-rate comparing with known ternary crystal structures. Moreover, we employ enhanced molecular dynamics (MD) simulation to further validate the stability of the ternary complex to incorporate the protein and PROTAC flexibility. We are actively developing deep-learning based methods to build the PROTAC database and assist the PROTAC design.
Q3：Currently, CADD is closely integrated with chemistry, what do you think about the synergy between the two? Can you share more information on how this is reflected in Viva’s platform?
Dr. Qian: The dynamics of computational chemistry with medicinal chemistry is collaborative and iterative. We propose plausible models to help understand the structure-activity relationship and make perspective predictions to guide the design and prioritize the compounds in the pipeline. Medicinal chemists in turn provide their feedback on how we should approach the design ideas and optimization of drug candidates accordingly. The inputs from both sides are highly valued with the progress of the drug discovery project. The CADD group at Viva work closely with internal and external medicinal chemists. We have been and are witnessing success compound design cases on multiple challenging drug discovery projects. We are incorporating more advanced CADD tools such as free energy perturbation (FEP) to the lead optimization stage that can further assist the decision-making with our medicinal chemistry colleagues.
Q4: In the past 20 years, with the rapid development of theoretical studies on receptor-ligand interactions and CADD methods, research on free energy prediction models has received increasing attention. Could you please share more details about Viva’s FEP platform and what are its unique features compared to external FEP platforms?
Dr. Qian: We've established our own FEP platform from scratch and hosted in on our own high performance computing system. It's a combination of user-friendly interface, fully-automated process, all-aspect analysis, and golden-standard accuracy. Since the platform is built in-house with a comprehensive understanding of the algorithm, we have access to all of the parameters and are capable of optimizing the FEP calculation conditions with first-handed experience. Currently the performance of the FEP platform is comparable to the top-tier commercialized software, according to a benchmark comparison of eight different protein systems. Equipped with the computing power at Viva, relative FEP calculations are now routine practices for the majority of the drug discovery projects.
Copyright © Viva Biotech All Rights Reserved. 沪ICP备19036061号
- About Us
-
CRO Services
BackCRO ServicesDiscovery Biology Services
- CDMO Services
- EFS Business
- News
- Careers
- Investor Relations
- Contact Us