Single unified syntax
One format for every genomic file — unaligned and aligned reads, references, and reports — so a whole analysis lives in a single interoperable file instead of a scattered toolchain.
The MPEG‑G platform
ISO/IEC 23092 MPEG‑G for genomic data, built for the realities of clinical workflows, multi-infrastructure deployment, and lossless compression at scale.
Provenance · 30 years of standards
The group behind MP3, HEVC and streaming video
MPEG — the Moving Picture Experts Group — operates under ISO and has spent three decades producing the compression standards behind MP3, DVD, Blu-ray, H.265/HEVC, Netflix and HBO. With 1700 delegates from 40+ countries, it specifies standards under non-discriminatory ISO/IEC licensing. MPEG‑G is the joint project of MPEG and ISO/TC 276 (Biotechnology) that brings the same rigour to genomic data.
- MP3
- DVD
- Blu-ray
- H.265 / HEVC
- Streaming video
- MPEG‑G
The benefits
What it means for your data
Built-in security
Privacy is part of the syntax: encrypt genomic data natively inside the file and grant access region by region, so sensitive sequences stay protected without bolt-on tooling.
Efficient storage
Around 83% smaller than BAM (≈6×), losslessly — up to ~90% on lighter data. Whole-exome and whole-genome datasets cost far less to store and move — without dropping a single base call.
Fast selective access
Read any region without unpacking the whole file. MPEG‑G decodes individual bases directly, cutting the latency before an analysis can even start.
Capabilities
What MPEG‑G makes possible
Lossless compression
Around 83% smaller than BAM (≈6×) on full whole-genome data, losslessly — up to ~90% (~10×) on lighter data. The latest transcoder reaches the same size 2.2× faster, using 1.9× less CPU.
Selective access
Retrieve any region of the genome in sub-second time, without decompressing the whole file.
Built-in integrity
Digital signatures, provenance chains, and consent metadata are part of the syntax — not an afterthought.
Multi-infrastructure
One pipeline runs unchanged on cloud, edge, or directly on a patient's phone.
Open standard
Published as ISO/IEC 23092 — the only open international standard for genomic data, developed under ISO with documented amendments and no lock-in.
Interoperable
Native MPEG‑G readers from GenomSys plus a growing ecosystem of partners and toolchains.
Inside the standard
One standard, six parts
ISO/IEC 23092 is specified in six parts — together they cover the full path from raw sequencer output to protected, queryable genomic data.
File & Transport Format
The technology to transport and access data — including progressive streaming over any network.
Genomic Information Representation
The compressed representation at the core of the standard.
APIs & Metadata
Standard interfaces with DNA applications and legacy formats, plus metadata for content protection and annotation.
Reference Software
An open-source normative decoder to support and guide implementers of MPEG‑G.
Conformance
The methodology to test compliance — an MPEG‑G–labelled device passes 100+ documented conformance tests.
Annotations
A unified compressed format for high-level analysis.
Profiles
A prospective seventh part defining profiles — curated subsets of the standard's tools tuned to specific application domains, so implementations can interoperate at a known, testable capability level.
Graph genome
A possible eighth part is on the horizon — a standardized graph-genome (pan-genome) representation — explored together with GenoGra.
Open ecosystem
Backed by an open-source community
The reference software is open source, and GENIE (GENomic Information Encoding) is a joint, collaborative effort to produce an MPEG‑G–compliant open-source encoder, with MPEG‑G integrated into GATK. Contributors span Stanford, the University of Illinois, UPC BarcelonaTech, Ghent University, Leibniz University Hannover and the Barcelona Supercomputing Center — alongside research partners including the Swiss Institute of Bioinformatics and the US National Cancer Institute.