ScienceAugust 10, 2023

AI in Chemistry

AI has many different applications in chemistry, with various concepts and databases, which AI handles being far too complex for even the most experienced chemists. To tackle these problems, significant advancements have been made by AI in many different, if not all, fields in chemistry.
Avatar Matteo MEUNIER

AI has many different applications in chemistry, with various concepts and databases, which AI handles being far too complex for even the most experienced chemists. To tackle these problems, significant advancements have been made by AI in many different, if not all, fields in chemistry. During my internship at Dassault Systems, BIOVIA. I have looked at research studies exploring these AI-driven fields, how they have progressed through time, and their future directions.

The relevance of AI in chemistry has recently had a sharp incline, with its importance dramatically increasing since 2015, with significant steps in Analytical Chemistry and Biochemistry. The relevant fields have gained popularity within journal publications and patents on the effects of AI in their respective disciplines, and it is predicted that even more will come in the following years.

Publication trends of AI in specific research areas from 2000 to 2020 [1]
  • A – Increase in AI-related journal publications
  • B – Increase in AI-related patent publications


The growth of AI publications and research discovery has become more relevant for a various reasons. Powerhouse countries increased their research efforts with a $93.5 billion investment just in 2021; this indicates that the real-world applications of this research will increase efficiency in many disciplines of chemistry, industry, and even everyday life. Thanks to recent advances in technology, high computational computational power has worked to tackle problems and AI research, and even made the systems more cost-effective, increasing accessibility to research for a greater number of companies. The understanding of both the potential and ability to make efficient research has also increased.

AI Empowering Researchers

AI-wielding software has also become more widespread in recent times, with the release of open-source AI machine learning framework projects such as TensorFlow (2015) and PyTorch (2016) coinciding with the AI boom.

During my internship, among researching AI publications and applications, I tried  BIOVIA Materials Studio, and it is impressive to see how accessible these powerful research tools are to people. BIOVIA Materials Studio offers a range of AI-based programs which provide aid to fields such as prediction of polymer properties (Synthia) and statistical analysis algorithms (QSAR) in drug discovery and optimization and prediction of materials properties. The relative accessibility of these programs using advanced machine learning algorithms, such as the genetic functions algorithms or neural networks, is expected to have a large impact on research. This means anyone from an undergraduate student to research professors will have a chance to make an impact using these effective methods of research

Example of a study table with similar molecules and their properties predicted using MS QSAR [2]

Drug design is a field where its importance to society is very well known, as we rely on drugs for highly important things such as medication and food. In my research, I identified a trend that saw media outlets mainly discussing this field, with specifically its potential for raising the quality of life and improving its efficiency and progress being a main topic for discourse. Opportunities in drug design for AI and machine learning exist in automation processes and overall improvement in the processes we use today. These improvements will accelerate these essential developments, make an extremely time-consuming and expensive process more viable for chemists and companies, and have many real-world positive implications. [3]

Another application example of AI in chemistry is in retrosynthesis, where complex molecules and drugs can be traced back to their industry products from an expansive database of synthesis reactions. Where chemists typically have to spend countless hours filtering out possible reaction pathways, commercial retrosynthesis AI software can automatically sort out unfeasible and useless pathways and give commercially viable ones. The algorithm runs through reactions that have harmful side products, expensive starting compounds, or side reactions that mess with protected functional groups. This filtering would otherwise be an extremely laborious and time-consuming process, something completely skipped by using such AI algorithms.

The Future of AI in Chemistry

The future of AI is endless with opportunities for improvement, and it can be difficult to locate a ceiling for how productive some of these disciplines in chemistry can become. This certain and undeniable rapid growth of AI and the implications of a perfect fully automated AI rendition of chemistry can even put into question the role of chemists in this future. Will it remove the need for chemists entirely? Will it end up losing jobs instead of creating them? Thankfully, researchers put these dystopian futures extremely far if not improbable. The growth for now is rapid, but not uncontrollable, and there is no doubt these developments will continue to be seen far into the future of chemistry research, as its potential in the real world is slowly filled.


  1. Baum, Zachary J., et al. “Artificial intelligence in chemistry: current trends and future directions.” Journal of Chemical Information and Modeling 61.7 (2021): 3197-3212.

Stay up to date

Receive monthly updates on content you won’t want to miss


Register here to receive a monthly update on our newest content.