Rustims: An Open-Source Framework for Rapid Development and Processing of timsTOF Data-Dependent Acquisition Data

TL;DR

Rustims is an open-source framework for processing timsTOF mass spectrometry data, combining fast Rust code with a user-friendly Python interface.

Contribution

Rustims introduces a dual-language, open-source framework for timsTOF DDA-PASEF data processing with integration of third-party tools.

Findings

Rustims supports both tryptic proteomics and nontryptic immunopeptidomics data processing.

Benchmark comparisons show Rustims performs well against existing tools like FragPipe and PEAKS.

The framework includes a pipeline with rescoring and integration of tools like Prosit and an extended ion mobility model.

Abstract

Mass spectrometry
is essential for analyzing and quantifying biological
samples. The timsTOF platform is a prominent commercial tool for this
purpose, particularly in bottom-up acquisition scenarios. The additional
ion mobility dimension requires more complex data processing, yet
most current software solutions for timsTOF raw data are proprietary
or closed-source, limiting integration into custom workflows. We introduce
rustims, a framework implementing a flexible toolbox designed for
processing timsTOF raw data, currently focusing on data-dependent
acquisition (DDA-PASEF). The framework employs a dual-language approach,
combining efficient, multithreaded Rust code with an easy-to-use Python
interface. This allows for implementations that are fast, intuitive,
and easy to integrate. With imspy as its main Python scripting interface
and sagepy for Sage search engine bindings, rustims enables fast,
integrable, and intuitive processing. We demonstrate its capabilities
with a pipeline for DDA-PASEF data including rescoring and integration
of third-party tools like the Prosit intensity predictor and an extended
ion mobility model. This pipeline supports tryptic proteomics and
nontryptic immunopeptidomics data, with benchmark comparisons to FragPipe
and PEAKS. Rustims is available on GitHub under the MIT license, with
installation packages for multiple platforms on PyPi and all analysis
scripts accessible via Zenodo.

Figures