NMRPy: a novel NMR scripting system to implement artificial intelligence and advanced applications

TL;DR

NMRPy is a new scripting system that integrates Java and Python to enable flexible, advanced NMR data analysis and instrument control, supporting emerging AI and statistical techniques.

Contribution

It introduces a novel bridge-based NMR scripting environment combining Java and Python, facilitating complex analysis and AI integration for NMR research.

Findings

Supports Java and Python libraries for NMR analysis

Enables easy integration of deep learning and statistical methods

Provides a user-friendly, open-source environment

Abstract

Background: Software is an important windows to offer a variety of complex instrument control and data processing for nuclear magnetic resonance (NMR) spectrometer. NMR software should allow researchers to flexibly implement various functionality according to the requirement of applications. Scripting system can offer an open environment for NMR users to write custom programs with basic libraries. Emerging technologies, especially multivariate statistical analysis and artificial intelligence, have been successfully applied to NMR applications such as metabolomics and biomacromolecules. Scripting system should support more complex NMR libraries, which will enable the emerging technologies to be easily implemented in the scripting environment. Result: Here, a novel NMR scripting system named "NMRPy" is introduced. In the scripting system, both Java based NMR methods and original CPython based libraries are supported. A module was built as a bridge to integrate the runtime environment of Java and CPython. It works as an extension in CPython environment, as well as interacts with Java part by Java Native Interface. Leveraging the bridge, Java based instrument control and data processing methods can be called as a CPython style. Compared with traditional scripting system, NMRPy is easier for NMR researchers to develop complex functionality with fast numerical computation, multivariate statistical analysis, deep learning etc. Non-uniform sampling and protein structure prediction methods based on deep learning can be conveniently integrated into NMRPy. Conclusion: NMRPy offers a user-friendly environment to implement custom functionality leveraging its powerful basic NMR and rich CPython libraries. NMR applications with emerging technologies can be easily integrated. The scripting system is free of charge and can be downloaded by visiting http://www.spinstudioj.net/nmrpy.