Reaxys Predictive Retrosynthesis

Boost success rates in synthesis route design and prediction

Retrosynthetic analysis once relied solely on a chemist’s expertise and time. The award-winning Reaxys Predictive Retrosynthesis changes that paradigm. It augments your knowledge with robust predicted routes by applying AI to the world’s largest database of chemical reactions. Get faster answers for a broader range of molecules with Reaxys Predictive Retrosynthesis.

GHP Life Sciences Award winner

GHP chose Reaxys Predictive Retrosynthesis as its 2020 Life Sciences Award winner. They state: “By combining Reaxys’ content with cutting-edge AI and machine learning technologies developed by PAI, the team is able to offer solutions unparalleled by competitors in the market.”

2019 International Life Sciences awards logo

What are the advantages of Reaxys Predictive Retrosynthesis?

Reaxys Predictive Retrosynthesis is a best-in-class innovation in retrosynthesis for medicinal and synthetic chemistry. Getting from a desired novel molecule to an optimal synthesis route used to take considerable time, effort and expertise. With Reaxys Predictive Retrosynthesis, you will see improvements in:

Route design

Retrosynthesis picto illustration

Find routes to synthesize novel compounds and new, optimized routes for existing chemical compounds

Decision making

Background data picto illustration

Get the background literature that informed predicted synthesis routes and experimental procedures to decide what chemical compound to make and how to make it


Faster prediction picto illustration

Predict synthesis routes much faster than with other computer-aided retrosynthesis solutions; routes can be generated for novel compounds in 10 minutes

Cost control

Commercial availability and pricing picto illustration

Determine commercial availability and pricing of starting materials as synthesis routes always terminate in purchasable starting materials

You can also integrate your own reaction data, building blocks or preferred vendors for more relevant results.

Elsevier and Pending.AI (PAI) collaborated to develop Reaxys Predictive Retrosynthesis based on foundational work published in Nature. We applied AI and machine learning to high-quality Reaxys data. And we double-blind tested predicted synthetic routes — proven to be scientifically sound and robust — with 45 organic chemists from around the world. The result is a solution that empowers chemists in:

  • Drug discovery
  • Agro-chemicals
  • Fine chemicals
  • Other chemical compound development

Reaxys Predictive Retrosynthesis uniquely leverages:

High-quality reaction data in Reaxys to train the predictive model — Deep learning technology has been trained on over 15 million reactions to auto-derive more than 400,000 reaction rules. These algorithmically extracted (versus hand-coded) rules are used to design and predict synthesis routes for known and novel molecules.

Three deep neural networks to auto-derive and apply reaction rules — The solution uses a model that incorporates deep neural networks trained on Reaxys data. The results are found using a Monte Carlo tree search approach to quickly discover promising candidate routes.

Self-learning from an ever-growing chemistry knowledgebase — The predictive model continuously improves as it gains intelligence from new data within Reaxys and the customer’s in-house research. This eliminates the negative impact of data siloes and unlocks learnings from successful and failed experiments.

How to use Reaxys Predictive Retrosynthesis

Medicinal and synthetic chemists need to reduce synthesis failures and validate proof of concept at the earliest stage possible. You need access to the most reliable and comprehensive reaction data, including your own. And you want to easily select optimal routes and manage your synthesis projects effectively.

Enter the target molecule and select parameters

Enter target molecule and select parameters screenshot

Choose among parameters for diversity and speed, and from libraries of building blocks. You can also add information from preferred vendors.

Get all published and predicted routes for a molecule in one view

Publish and predict molecule routes screenshot

Compare the number of steps per route, the route linearity and confidence ranking. Export routes in multiple formats to share and add to electronic lab notebooks.

Find purchasable or in-house starting materials

Purchasable or in-house starting materials screenshot

Integrate your own library of starting materials, stockroom compounds or proprietary molecules.

Get access to Reaxys Predictive Retrosynthesis

Complete the following form and we will contact you to discuss how Reaxys Predictive Retrosynthesis can support your chemical research and development.