Neuroline: an integrated modular pipeline for FAIR multimodal neurophysiology data processing

Abstract

The adoption of FAIR principles for code and data sharing in systems neuroscience is hindered by the size and diversity of datasets, the multitude of data formats and the complexity of analysis workflows. These challenges have led to the development of various data standards and modality-specific processing and analysis tools. However, it remains difficult to integrate the diverse processed data streams originating from state-of-the-art neurophysiology experiments into a common format, which is crucial for cross-modal analyses and open, FAIR sharing of data. To address this, we developed Neuroline, a Python package for processing neurophysiology data (electrophysiology, optical imaging, and behaviour) that combines multimodal integration, modularity, and standardised output within a single pipeline. It leverages established packages and consolidates the resulting processed data into the standardised Neurodata Without Borders (NWB) format. Neuroline provides a unified interface while preserving flexibility and extensibility. Here we describe the implementation and challenges of a pipeline capable of integrating different types of neuroscientific data and packages. Finally, we address how Neuroline contributes to the sustainability and accessibility of neuroscience data and software.