ALPAR: automated learning pipeline for antimicrobial resistance

Area of Science:

• Genomics
• Bioinformatics
• Machine Learning

Background:

• Antimicrobial resistance (AMR) research is rapidly advancing, driven by genomic sequencing and computational power.
• Challenges exist in data preparation and bioinformatic analysis for AMR research, particularly for new researchers.
• A need exists for streamlined, reproducible tools to analyze AMR genomic data.

Purpose of the Study:

• To introduce ALPAR, an Automated Learning Pipeline for Antimicrobial Resistance.
• To provide a comprehensive, automated workflow for AMR data analysis from raw genomic data to results interpretation.
• To simplify the process of generating machine learning-ready data and performing analyses like GWAS.

Main Methods:

• ALPAR processes raw genomic data (FASTA files) through a reproducible pipeline.
• It integrates established bioinformatics tools for automated data preparation and analysis.
• The pipeline supports machine learning model training and genome-wide association studies (GWAS).

Main Results:

• ALPAR generates machine learning-ready data tables from genomic inputs.
• It facilitates the training of machine learning models for AMR prediction.
• The tool successfully competed in benchmarks, winning a 2024 challenge and placing third in a 2025 challenge.

Conclusions:

• ALPAR addresses the complexity of AMR data analysis, offering a simplified, automated solution.
• The tool enhances reproducibility and accessibility for researchers in the AMR field.
• ALPAR's performance in competitive benchmarks demonstrates its effectiveness and utility.