Open, Reproducible Calculation of Assembly Indices

TL;DR

This paper introduces assembly-theory, a Rust software package that efficiently calculates assembly indices for molecular structures, supporting research in complexity and selection across chemical systems.

Contribution

It provides an extensible, high-performance implementation of assembly index algorithms with benchmarking and integration tools for the scientific community.

Findings

Efficient computation of assembly indices for covalent molecules.

Benchmarking framework for algorithmic improvements.

Python and RDKit integration for broader usability.

Abstract

We present assembly-theory, a Rust package for computing assembly indices of covalently bonded molecular structures. This is a key complexity measure of assembly theory, a recent theoretical framework quantifying selection across diverse systems, most importantly chemistry. assembly-theory is designed for researchers and practitioners alike, providing (i) extensible, high-performance implementations of assembly index calculation algorithms, (ii) comprehensive benchmarks against which current and future algorithmic improvements can be tested, and (iii) Python bindings and RDKit-compatible data loaders to support integration with existing computational pipelines.