SynKit#
SynKit is a graph-based Python toolkit for reaction informatics: canonicalization, matching, rule manipulation, and lightweight chemical reaction network (CRN) exploration. It is designed to be modular, extensible, and friendly for both research and production workflows.
Version 1.4.0 highlights#
LSG stores sigma/pi orders, lone pairs, radicals, valence electrons, and charge recomputation metadata for the pure-Python reactor.
Native networkx functional-group detection now replaces the previous
external FG helper and returns graph-indexed labels.
Mechanistic Transition Graphs keep temporal bond histories and can be projected back to ordered ITS steps or composed transformations.
Modern drawers cover molecule graphs, reaction panels, ITS-only views, Lewis-state labels, and compact MTG timelines.
Documentation quick links#
Install, quick examples, and basic usage.
CRN building, hypergraph conversions, canonicalization utilities.
Readers/writers and format conversion recipes.
Reaction validation, AAM checks, balance and fingerprinting.
ITS/MTG construction, WL hashing, and cluster/search primitives.
Molecule, reaction, ITS, and diagnostic graph rendering.
Auto-generated API docs and class references.
Citation#
Please cite SynKit when used in published work. Prefer the archived DOI for a specific release and include the Git commit/tag for reproducibility.
@article{phan2025synkit,
title = {SynKit: A Graph-Based Python Framework for Rule-Based Reaction Modeling and Analysis},
author = {Phan, Tieu-Long and others},
year = {2025},
doi = {10.5281/zenodo.15269901}
}
Contributions & support#
Report issues or feature requests on GitHub: TieuLongPhan/SynKit#issues To contribute: fork, create a topic branch, add tests, and open a PR against main. For API changes update docs and tests. For contribution enquiries or collaborations: tieu@bioinf.uni-leipzig.de
Acknowledgments#
Funded by the European Union Horizon Europe Doctoral Network (Marie-Skłodowska-Curie, grant No. 101072930 — TACsy).