AMR Interpretation Engine Brigham and Women's Hospital
winget install --id=WHONET.AMRIE -e This software was created by the WHONET development team (https://whonet.org) to facilitate the interpretation of antibiotic measurements according to the various supported guidelines available for this purpose, and published by those groups (CLSI, EUCAST, etc.). The system can be integrated with your project via the Interpretation Engine library, or it can be used directly with either the command line interface or interactive interface projects. The interactive interface is designed primarily to allow you to exercise the various functions of the system, but it can also allow you to process an input file into an output file with the interpretations. The command-line interface facilitates easy incorporation with a 3rd-party system since the developer does not need to know how to use the various library functions associated. With that said, the CLI project is very small, so it should serve as a window into how to accomplish the basic needs if direct library integration was preferred. The breakpoints and other tables will be kept up-to-date over time (Interpretaion Engine\Resources), and are useful in their own right, even if the software implementation of the engine is not for your purposes. For example, there are several groups who only utilize our tables. This repository now serves as the official source for these WHONET-related resources. To process a complete data file, the input file must use WHONET naming conventions for the data columns. Sample data and configuration files are also provided in the Interpretation Engine\Resources\ folder. We also provide SQL queries which can demonstrate certain aspects of the system, but which are not used directly by the code here. We have chosen to implement these function as pure C# code to assist others if they choose to make their own implementations, and also to eliminate any dependence on a certain database technology. Because the application only uses C# with a recent .NET version, it should be possible to build the library code for many platforms.
AMR Interpretation Engine is a software tool designed to facilitate the interpretation of antibiotic susceptibility measurements according to established guidelines, such as those published by CLSI (Clinical and Laboratory Standards Institute) and EUCAST (European Committee on Antimicrobial Susceptibility Testing). This system was developed by the WHONET team to provide accurate and reliable interpretations for microbiology professionals.
Key Features:
- Library Integration: The software provides a library that can be directly integrated into third-party projects, enabling seamless interpretation of antibiotic data without requiring knowledge of the underlying functions.
- Command-Line Interface (CLI): A lightweight CLI tool allows easy incorporation into workflows or scripts, making it ideal for automation and batch processing.
- Interactive Interface: Designed for hands-on exploration, this interface supports ad hoc queries and provides a user-friendly way to generate interpretations for entire data files.
- Updated Breakpoints: The system maintains up-to-date breakpoints and tables, ensuring accurate interpretation based on the latest guidelines.
- WHONET Naming Conventions: Input files must adhere to WHONET naming standards for consistent processing of measurements, organisms, and antibiotics.
- Cross-Platform Support: Built with .NET 8, the application is compatible with multiple platforms, offering flexibility for diverse environments.
Audience & Benefit: Ideal for microbiologists, researchers, and clinical laboratories involved in antimicrobial susceptibility testing, this tool streamlines interpretation processes and ensures compliance with international guidelines. By providing accurate and efficient interpretations, it supports informed decision-making in treatment plans and helps monitor trends in antimicrobial resistance.
The software can be installed on Windows using winget install WHONET.AMRIE.
README
AMRIE
Antimicrobial Resistance Test Interpretation Engine
This software was created by the WHONET development team to facilitate the interpretation of antibiotic measurements according to the published guidelines. For more information about the guidelines, please visit their websites: CLSI, EUCAST, SFM
You can install this software on Windows using the command winget install WHONET.AMRIE or by downloading the latest release from GitHub.
There are three different sets of needs that the solution aims to support:
- Interactive use
- Command line use
- Library integration
The library code is written in .NET 8 with no additional package dependencies. It can be directly integrated with other projects on any platform which supports .NET 8. The interactive and command line interfaces exercise various interpretation features.
The interactive system may be used to generate the interpretations for an entire data file, or explore ad hoc interpretations and view filtered resources.
Ad hoc queries allow you to select from lists of organisms and antibiotics, and provide an associated measurement to interpret.
To generate interpretations for an entire data file using the interface or command line application, the file must use WHONET naming conventions for the
data columns, measurements, organism codes, etc. Sample data and configuration files are provided in the installation package's Resources folder.
The command line interface is another option which can be used to generate interpretations without requiring knowledge of or integration with the library. Instead, the command line application can be called as needed (from another application, from a batch script, etc.). It should be possible to build the command line interface for platforms other than Windows with minimal changes.
We also provide SQL queries that demonstrate certain concepts, but which are not used by the code. These concepts are implemented in C# to assist others if they choose to make their own implementations, and also to eliminate any dependence on a certain database technology which could restrict cross-platform options.
To improve processing perfomance without the help of the database, the system has automatic breakpoint caching built in, and includes a way to "preheat" the breakpoint cache to maximize interpretation performance if additional information about the input data set can be provided ahead of time.