Talos: Scaling rare disease diagnosis with automated, iterative genomic reanalysis

2026-06-24 · Source: Microsoft Research · Field: Health & Wellbeing — Clinical Care & Medical Practice, Medical Devices & Health Technology, Artificial Intelligence & Machine Learning · Depth: Advanced, medium

Summary

Talos is an open-source tool designed for automated, iterative reanalysis of genomic data, specifically targeting rare disease diagnosis. Developed through a collaboration including the Centre for Population Genomics and Microsoft, Talos efficiently re-examines stored sequencing data as scientific knowledge evolves, flagging variants with newly actionable evidence. Across a validation set of nearly 1,100 patients, it recovered 90% of in-scope diagnoses while flagging only 1.3 candidate variants per patient for expert review, demonstrating a low false-positive rate crucial for scalability. When deployed on a prospective cohort of almost 5,000 undiagnosed patients, Talos delivered 241 new diagnoses, representing a 5.1% additional yield. This system reduced the average time between supporting evidence becoming public and diagnosis to just 32 days. Monthly iterative cycles required analysts to review only one new variant per 200 patients, proving systematic reanalysis can be sustained.

Key takeaway

For clinical genomics labs or health systems aiming to improve rare disease diagnostic rates, implementing automated genomic reanalysis like Talos is crucial. You can significantly increase diagnostic yield by continuously re-evaluating patient data against evolving scientific knowledge, reducing the time to diagnosis. Consider deploying open-source solutions in cloud environments to manage the computational load and minimize manual review, making frequent reanalysis sustainable and cost-effective.

Key insights

Talos automates genomic reanalysis for rare diseases, significantly increasing diagnostic yield by utilizing evolving scientific knowledge.

Principles

• Optimize for specificity to reduce human review time.
• Continuously update variant knowledge from public resources.
• Focus on newly actionable evidence for iterative cycles.

Method

Talos re-interprets existing variant calls against PanelApp Australia and ClinVar, applying a prioritization algorithm. It refines candidates using family structure and phenotype, then reports only newly actionable variants.

In practice

• Integrate PanelApp Australia and ClinVar for current data.
• Prioritize variants based on ACMG/AMP criteria.
• Deploy in cloud environments like Azure for scalability.

Topics

• Genomic Reanalysis
• Rare Disease Diagnosis
• Variant Prioritization
• ClinVar
• PanelApp Australia
• Open-Source Tools

Code references

• populationgenomics/talos

Best for: Research Scientist, AI Scientist, MLOps Engineer

Related on AIssential

• Mitosis Detection in the Wild: Multi-Tumor and Context-Aware Generalization in the MIDOG 2025 Challenge — "In the wild" mitosis detection demands models robust to diverse biological and contextual variance.
• Accelerating discovery of liver disease mechanisms — AI-driven literature analysis can uncover novel biological hypotheses and candidate therapies for complex diseases.
• Interpretable machine learning and signal processing for automated reading and quality control of lateral flow tests for schistosomiasis — An automated pipeline combining deep learning and signal processing accurately classifies schistosomiasis lateral flow tests.
• Databricks for Good and Virtue Foundation: Partnering to Connect Medical Volunteers to Critical Health Services in 72 Countries — Production-grade LLM pipelines and unified data platforms can transform disparate web data into actionable global health insights.
• Active Evidence-Seeking and Diagnostic Reasoning in Large Language Models for Clinical Decision Support — LLMs exhibit a substantial performance gap in active, multi-turn clinical diagnosis compared to static, full-context evaluations.
• Spotlight Pathology has raised £1.4M to catch blood cancer sooner — AI-powered image analysis can significantly improve the speed and consistency of blood cancer diagnosis.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Microsoft Research.