Do you ever… 

… wonder how SYNTHIA™ works out which reactions are suggested in your pathways?

 

There are two types of Reaction Databases in SYNTHIA™:

Rule-based: Reaction rules are based on transformations you would find in organic chemistry textbooks or in the literature. These rules are used to look at possible disconnections for your target molecule and suggest practical reactions to build up the molecule. They are written in a general way, so that they can be applied to any molecule, including novel ones.

Published reactions: Reactions that have been published in the literature.

You can select which databases you prefer SYNTHIA™ to use in the Analysis Configuration setting page, under the Analysis Preferences tab in the Advanced section.


SYNTHIA™ is built on our Expert-Coded rules Database, which consists of over 100,000 reaction rules. These rules are based on transformations you would find in organic chemistry textbooks or in the literature and are written in a general way, so that they can be applied to any molecules, including novel ones. They consider full stereo- and regio-chemistry, and include protection information, reactivity conflicts, literature sources, and reaction conditions. 

There is an additional subset of expert-coded rules based uniquely on enzymatic reactions, that can suggest enzymatic reactions as a green alternative to traditional chemistry.

By applying these rules to potential retrosynthetic disconnections, SYNTHIA™ suggests reactions that are possible based on the chemistry that we know, and compatible with the rest of the molecule structure.


There are five additional Machine-Extracted Rules databases, which cover specialized chemistries. They may be used to supplement the expert-coded rules database, particularly if your structure contains stereochemistry, complex aromatic systems, or unusual heterocycles.

These rules were automatically extracted from the literature and may not contain all possible reaction condition incompatibilities, reactivity conflicts or protection requirements.

  • Uncommon chemistry: Reaction rules for uncommon transformations that have a minimum of 10 examples.
  • Rare chemistry: Reaction rules for rare transformations that have a minimum of 3 examples.
  • Stereoselective chemistry: Reaction rules for stereoselective transformations.
  • Heterocyclic chemistry: Reaction rules for the synthesis of aromatic heterocycles.
  • Rhea reactions rules: Reaction rules extracted from the Rhea database of chemical and transport reactions of biological interest, which includes enzymatic reactions and transport reactions as well as reactions that occur spontaneously in biological systems, such as metabolic reactions.

The machine extracted rules may show you some unique chemistry, including deprotection reactions or unusual heterocycle formations. However they do not contain illustrative references.


In addition to reaction rules databases, SYNTHIA™ also includes several databases of reactions that have been published in the literature.

  • SPRESI by DeepMatter covers organic synthesis literature from 1974 to 2014, and contains 4.6 million chemical reactions abstracted from 700,000 references
    >> SPRESI by DeepMatter
  • Pistachio by NextMove includes over 10 million reactions sourced from patents filed with the United States Patent and Trademark Office (USPTO), the European Patent Office (EPO), and the World Intellectual Property Organization (WIPO), up to the year 2024.
  • Rhea is an expert-curated database of biochemical reactions, that includes enzymatic reactions and transport reactions as well as reactions that occur spontaneously in biological systems, such as metabolic reactions. 
    >> Rhea
  • Science of synthesis provides a critical review of the synthetic methodology developed from the early 1800s to-date for the entire field of organic and organometallic chemistry. 
    >> Science of Synthesis


By default, all expert-coded rules are enabled, along with the machine-extracted rules for Heterocyclic Chemistry and the SPRESI, Pistachio, and Science of Synthesis databases.

Machine-extracted rules are meant to complement the expert-coded rules database with additional chemistry that may be missing. If they are enabled, they may significantly increase the analysis time to completion, as they add additional rules for Synthia to consider for each disconnection. For this reason, it is best to use them sparingly.

You can enable Stereoselective Chemistry when you are looking how to form a chiral bond for your molecule, or Heterocyclic Chemistry if your target molecule contains an unusual heterocycle, or both if your target molecule is a natural compound. Use Uncommon Chemistry and Rare Chemistry when you are not finding results or want to find more diverse pathways.

You can disable the Pistachio  database if you are not interested to see patent reactions in your results, or disable all published reaction databases to find only predicted pathways.

 

In your results, you can identified reactions coming from Published Reaction databases or Machine-Extracted Rules databases from their name in Path View. Published reactions also display a book tag with three stars and a ‘100’ badge.

In Graph View, the reactions diamonds are colored in blue if they come from Machine-Extracted Rules databases or green if they come from Published Reaction databases.


To easily find pathways that contain published reactions or reactions suggested from machine-extracted rules, you can filter your results using 'published' or 'machine'  as keywords.