Overview
AWS HealthOmics is used in a reference architecture for processing public bacterial sequencing data as part of antimicrobial-resistance surveillance. It runs a managed Nextflow workflow, connecting read quality control, genome assembly, and resistance-gene detection in one genomics-processing pipeline.1
Use in sequencing workflows
In the described workflow, public sequence data is retrieved from the NCBI Sequence Read Archive through an ingestion process coordinated by Step Functions and an ECS/Fargate task. DynamoDB is used to avoid processing the same data more than once.1
The HealthOmics Nextflow pipeline applies fastp for quality control of sequencing reads and SKESA for de novo assembly. It then uses AMRFinderPlus to screen assembled contigs for known resistance genes.1 These stages turn raw sequence reads into results that can be normalized and queried downstream.1
Downstream data flow
A Lambda function normalizes pipeline output to the hAMRonization schema. AWS Glue combines the normalized results into Iceberg tables on Amazon S3 Tables, and Athena makes the resulting data available through SQL queries.1 In the demonstration, a Strands agent on Amazon Bedrock AgentCore uses Athena-backed tools to answer questions about detected resistance genes in plain language.1
The architecture is defined in AWS CDK and is intended as a demonstration that makes AMR data more accessible to researchers without specialized bioinformatics skills; it is not presented as a clinical system.1
Related Pages
- Antimicrobial Resistance Surveillance — AWS HealthOmics can run sequencing workflows that detect resistance genes, providing data and analysis relevant to antimicrobial resistance surveillance.