The neural network performs optical recognition of chemical compound structures and Markush structures from any documents such as patents, scientific articles, test protocols and dissertations. This simplifies and speeds up the process of collecting and analysing data for research, developing new compounds and writing literature reviews

PDF2SMILES

Automatic extraction of structural formulae from PDF documents

Cc1ccc(-c2nc(CN3C(=O)[C@@H](C)N(c4ccc(C)cc4)C3=O)cs2)cc1

You can upload your PDF documents, after which the system automatically starts the structure recognition process. This process takes place in the background and usually takes a few minutes. Recognition time depends on the size of the document. For more information, please read the following article

How it works

Saving to a dataset for further processing

Ability to edit recognised structures

Data export in formats: CSV, PNG

Characteristics

Reduction of time
manual processing of literature sources

The maximum possible document size is 250 MB

Video tutorial

If you would like to know more about Syntelly, leave a request and we will call you back

Ready to answer your questions