Illumina Unveils AI Genomic Analysis Tools with 99.4% Variant Detection Accuracy

Illumina Debuts Unified BioInsight Platform for Genomic Research

Illumina introduced its BioInsight Platform, a unified genomics software platform designed to streamline how researchers analyze genomic and multiomic data2

2

. The company created its BioInsight business unit one year ago to address challenges in converting large-scale genomic datasets into actionable insights2

2

. The platform consolidates Illumina's analysis applications, AI-driven capabilities, and curated data resources into a single sign-on environment, eliminating the need for researchers to assemble and manage separate informatics tools as their research questions evolve4

4

.

The BioInsight Platform operates on a usage-based credits model, replacing traditional annual subscriptions and allowing researchers to adopt capabilities most relevant to their work4

4

. The platform brings together Illumina's portfolio including DRAGEN, Emedgene, Connected Multiomics, BioInsight Platform Core, BaseSpace, and BioInsight Data2

2

. According to Rami Mehio, senior vice president and general manager of BioInsight at Illumina, the platform was "built on an incredible history of data, software, and AI innovation that Illumina has delivered in the industry for more than a decade"2

2

.

Three AI-Powered Genomic Analysis Tools Enter Early Access

At the American Society of Human Genetics annual meeting in Montréal, Illumina announced early access availability of three software capabilities on its BioInsight Platform: BioInsight Platform Workflow Orchestrator, DRAGEN Ultra, and BioInsight Platform Learning Cohorts1

1

. These tools are designed to automate genomic data analysis and accelerate genomic research workflows from sequencing data to discovery3

3

.

The Workflow Orchestrator connects to Illumina sequencers to automate genomic operations from sample tracking through analysis and quality control1

1

. A visual workflow builder allows laboratories to configure parameters and maintain sample identity across software tools1

1

. The software executes workflows consistently across tens of thousands of samples, enabling teams to scale operations without adding complexity while maintaining complete result lineage covering inputs, software versions, settings, and quality metrics3

3

.

DRAGEN Ultra Achieves 99.4% Accuracy in Variant Detection

DRAGEN Ultra extends genomic analysis beyond variant calling by incorporating AI models to prioritize genetic variants for investigation1

1

. The tool includes dnaAI, a variant prioritization model developed by Illumina's BioInsight AI Lab that predicts which variants are most likely to explain an observed phenotype3

3

. Research findings presented at the conference demonstrated that the model resulted in an 8% increase in disease associations across unsolved cases1

1

. The model ranked established disease-associated variants among its top four candidates in 99.4% of analyzed genomes1

1

.

DRAGEN Ultra is available on the BioInsight Platform and in customer premises, private clouds, and air-gapped environments1

1

. In addition to dnaAI, DRAGEN Ultra draws on Illumina models designed to address distinct classes of biological variation, including PrimateAI-3D, PromoterAI, and the newly announced SpliceAI2 for rare disease research3

3

. The technology is designed to support applications including rare disease research, population genomics, and multiomic studies3

3

.

DRAGEN v4.6 Expands Deployment Flexibility

The platform includes DRAGEN v4.6, an updated version of Illumina's secondary analysis software that expands deployment options2

2

. The new DRAGEN software mode allows organizations to run the analysis tool across supported cloud and standard CPU environments, expanding beyond previous versions2

2

. DRAGEN v4.6 delivers consistent analysis across supported compute environments, giving organizations flexibility to run DRAGEN where it best fits their needs while maintaining confidence in results4

4

.

The Global Parkinson's Genetics Program, a resource program of Aligning Science Across Parkinson's in partnership with The Michael J. Fox Foundation for Parkinson's Research, is using DRAGEN software mode to support approximately 170 active research projects investigating the genetic basis of Parkinson's disease2

2

. Faster processing and the ability to run analyses efficiently on a public cloud environment is helping the Global Parkinson's Genetics Program accelerate and share their large genomic cohort studies4

4

.

Learning Cohorts Tool Enables Dynamic Population Analysis

BioInsight Platform Learning Cohorts enables researchers to analyze genomic, multiomic, and phenotypic data across private sample collections1

1

. The tool allows users to build cohorts dynamically and investigate associations across datasets, moving beyond single sample evaluation into dynamic knowledge generation where every sample informs future discovery3

3

. Researchers can explore data across large private sample collections, compare individual observations with broader populations, and investigate associations that may not be visible within any one sample3

3

.

Researchers can begin with a scientific question, identify relevant populations from available datasets, and dynamically build cohorts for further analysis3

3

. For example, a researcher investigating a rare variant can identify subjects sharing similar genomic features or phenotypes, compare outcomes across groups, and explore associations that may warrant further study3

3

. The Learning Cohorts tool, paired with BioInsight Data, makes capabilities available to private researchers that were traditionally only available with large public cohorts3

3

.