Skip to main content
M
MIU
Explore Platform
Multi-Omics Integration Utility

MIU: Unified Multi-Omics Microbiome Analysis Platform

From amplicon to multi-omics integration — one reproducible, AI-assisted workflow for microbiome research.

Explore PlatformView Roadmap
Reproducibility
>80%
Taxonomy F1
>90%
Omics layers
6
MIUCore
Amplicon
Shotgun WGS
Metatranscriptomics
Metaproteomics
Metametabolomics
Omics Integration
The Problem

Fragmentation in Microbiome Analysis

Researchers stitch together disjoint tools across omics layers — sacrificing reproducibility, speed, and confidence.

ProblemImpact
Single-omics toolsDNA, RNA, protein, metabolites analyzed separately
Poor reproducibilityOnly 30% protein overlap across tools
High false positivesOnly 8% metabolite overlap across platforms
Steep learning curveQIIME2 takes months to master
No multi-omics integrationMOFA2 exists but disconnected
The Solution

Project MIU

MIU unifies microbiome workflows into a single platform with consensus-based and AI-assisted decision making.

Universal Parser + MIU-HostClean
Ingest any input format and strip host contamination automatically.
MIU-DomainMaster
Domain-aware taxonomic classification across bacteria, archaea, fungi, viruses.
Consensus Engine
Combine multiple tools to maximize reproducibility and reduce false positives.
MIU-INTEGRATE Suite
Native MOFA/DIABLO-style integration across all omics layers.
MIU-Capsule
One-command containerized workflows for full reproducibility.
Web Full-Featured Interface
Run, monitor, and share analyses from a unified browser experience.
Landscape

Comparison to Existing Software

Specialized tools each cover a slice. MIU covers the whole stack.

FeatureQIIME2VEBA2FragPipeXCMSMOFA2MIU
Amplicon (16S)
Shotgun Metagenomics
Metatranscriptomics
Metaproteomics
Metabolomics
Multi-Omics Integration
Long-read support
Web Full-featured
Coverage

The Gaps MIU Closes

CategoryKey GapsMIU Solution
Input & PreprocessingHost contamination, mixed formatsUniversal Parser + MIU-HostClean
Taxonomic ClassificationFalse positives, domain-specificMIU-DomainMaster + Consensus (Kraken2 + Centrifuge + MetaPhlAn4)
Metaproteomics30% tool overlapConsensus across 3+ engines (FragPipe, MaxQuant, MetaMorpheus)
Metametabolomics8% feature overlapConsensus across 4 platforms (XCMS, MS-DIAL, MZmine, iMet-Q)
Multi-OmicsNo MOFA/DIABLO integrationMIU-INTEGRATE suite
ReproducibilityNo built-in containerizationMIU-Capsule one-command
Methodology

Consensus and AI-Based Approach

MIU uses consensus methods and AI-assisted decision logic to reduce false positives, improve reproducibility, and automatically select the best analysis strategy based on the user's dataset.

DomainProblemMIU SolutionTarget Improvement
Metaproteomics30% overlapFragPipe + MaxQuant + MetaMorpheus
>80% reproducibility
Metametabolomics8% overlapXCMS + MS-DIAL + MZmine + iMet-Q
>50% reproducibility
TaxonomyDomain-specificKraken2 + Centrifuge + MetaPhlAn4
>90% F1 score
Auto-Denoiser Selection
Recommends DADA2 vs Deblur vs UNOISE based on user data.
False Positive Shield
Detects and removes false taxonomic assignments.
Adaptive Speed-Accuracy
Auto-selects fast vs accurate based on user needs.
Bioinformatics Workbench

A Full Bioinformatics Workbench, Built for Microbiomes

MIU brings the breadth of a modern molecular biology workbench — alignment, assembly, primers, BLAST, phylogenetics, annotation — into one reproducible, AI-assisted platform purpose-built for microbiome research.

Sequence Alignment
MIU-Align

Multiple sequence alignment for 16S, ITS, and full-length amplicons with MAFFT, MUSCLE, and Clustal consensus.

  • Pairwise & MSA
  • Codon-aware mode
  • Alignment quality scoring
De Novo Assembly
MIU-Assemble

Hybrid short- and long-read metagenomic assembly using SPAdes, Flye, and metaMDBG with auto-binning.

  • Hybrid assembly
  • MAG binning
  • Contig QC reports
Primer Design
MIU-Primer

Design and validate amplicon and qPCR primers with built-in in-silico PCR against SILVA and GTDB.

  • Tm / GC / hairpin checks
  • In-silico PCR
  • Coverage estimation
Sequence Search
MIU-BLAST

Federated BLAST and DIAMOND search across NCBI, SILVA, UNITE, GTDB, and custom reference databases.

  • BLASTn / BLASTp / DIAMOND
  • Custom DB upload
  • Hit ranking & taxonomy
Phylogenetics
MIU-Phylo

Build maximum-likelihood, neighbor-joining, and Bayesian phylogenies with one-click iTOL and Newick export.

  • IQ-TREE / RAxML / FastTree
  • Bootstrapping
  • iTOL & Newick export
Annotation
MIU-Annotate

Functional and taxonomic annotation with Prokka, eggNOG-mapper, KEGG, COG, CAZy, and Pfam.

  • Gene calling
  • Pathway mapping
  • Custom annotation tracks
Read Mapping
MIU-Map

Consensus read mapping across Bowtie2, BWA-MEM, and minimap2 with coverage and strain-resolved profiling.

  • Short & long reads
  • Coverage heatmaps
  • Strain inference
Variant Calling
MIU-Variants

Strain-level SNP and indel calling with bcftools, FreeBayes, and DeepVariant in a consensus workflow.

  • SNV / Indel / SV
  • Strain deconvolution
  • VCF & report export
Cloning & Synthetic Biology
MIU-Construct

Design constructs, validate restriction sites, and simulate cloning workflows for engineered microbes.

  • Restriction & Gibson
  • Plasmid simulation
  • Codon optimization
Extensible Workbench
MIU-Plugins

Extend MIU with Python and Rust plugins, custom panels, and reproducible workflow steps.

  • Python / Rust SDK
  • Custom panels
  • Workflow nodes
Collaboration
MIU-Workspace

Shared lab projects, role-based access, change history, and reproducible workflow snapshots.

  • Shared projects
  • Roles & audit log
  • Snapshots & forks
Visualization
MIU-Viewer

Interactive viewers for alignments, chromatograms, trees, contigs, and abundance heatmaps.

  • Alignment & trace viewer
  • Tree explorer
  • Abundance heatmaps
Alignment Viewer
16S V3–V4 · 5 samples
Sample_01
AGCTTAGCTAGGCTTAACGTTAGCTAAGCTAGGCTTAACGTAGCTA
Sample_02
AGCTTAGCTAGGCTTAACGTTAGCTAAGCTAGGCTTAACGTAGCTA
Sample_03
AGCTTAGCTAGACTTAACGTTAGCTAAGCTAGGCTTAACGTAGCTA
Sample_04
AGCTTAGCTAGGCTTAACGTTAGCTAAGCTAGGCTTAACGTAGCTA
Sample_05
AGCTTAGCTAGGCTTAACGTTAGCTAAGCTAGNCTTAACGTAGCTA
Consensus
AGCTTAGCTAGGCTTAACGTTAGCTAAGCTAGGCTTAACGTAGCTA
Identity: 98.6%Gaps: 0Variants: 2Quality: Q37 avg
Workbench vs. MIU
CapabilityClassic WorkbenchMIU
Sequence Alignment
De Novo Assembly
Primer Design + in-silico PCR
BLAST / DIAMOND Search
Phylogenetics
NGS Read Mapping & Variants
Annotation (Prokka / eggNOG)
Consensus Multi-Omics Integration
AI-assisted False Positive Shield
Containerized Reproducibility (MIU-Capsule)
Web-Native, Multi-Engine Architecture
Live Workbench

Interactive Modules — Alignment, Assembly, BLAST, Primers & Variants

A Geneious-class workbench, adapted for microbiome science. Zoom alignments, explore assemblies and gene maps, run BLAST against SILVA / GTDB / UNITE, design PCR primers, call variants, and export sequences and reports — all in MIU.

pos 1–45 / 106
V3 Primer
16S Conserved
Reference
AGCTTAGCTAGGCTTAACGTTAGCTAAGCTAGGCTTAACGTAGCT
read_1
AGCTTAGCTAGGCTTAACGTTAGCTAAGCTAGGCTTAACGTAGCT
read_2
AGCTTAGCTAGGCTTAACGTTAGCTAAGCTAGGCTTAACGTAGCT
read_3
AGCTTAGCTAGACTTAACGTTAGCTAAGCTAGGCTTAACGTAGCT
read_4
AGCTTAGCTAGGCTTAACGTTAGCTAAGCTAGGCTTAACGTAGCT
read_5
AGCTTAGCTAGGCTTAACGTTAGCTAAGCTAGGCTTAACGTAGCT
Coverage: 5×Identity: 98.6%Variant positions: 1Annotated regions: 4
Engineering

Architecture & Licensing

Module TypeLanguageOpen-SourceCommercial
Core InfrastructureRust + Python
MIT / BSD-3
Yes
High-Performance ComputeRust
MIT + Apache-2.0
Yes
Python/R/Julia APIMulti
LGPL-3.0
Yes
CLI / Pocket AppsRust + Python
GPL-3.0
Yes
Cloud ServicesRust + TypeScript
AGPL-3.0
Yes, SaaS
Enterprise FeaturesRust/Python
Proprietary
Yes, enterprise only
Computational Core

Multi-Engine Computational Core

MIU combines Python, AI, Rust, DUST, and Jupyter-based engines to deliver fast, reproducible, automated, and researcher-friendly microbiome analysis.

Python Engine

Handles bioinformatics pipelines, data preprocessing, statistical analysis, visualization, and integration with scientific Python libraries.

  • Microbiome data preprocessing
  • Pandas / NumPy / SciPy workflows
  • Scikit-learn model integration
  • Plotting and reporting
  • API bridge for bioinformatics tools
AI Engine

Provides intelligent workflow recommendation, false-positive detection, denoiser selection, model evaluation, and adaptive analysis decisions.

  • Auto-Denoiser Selection
  • False Positive Shield
  • Adaptive Speed-Accuracy
  • F1 score optimization
  • AI-assisted workflow recommendation
  • Multi-omics pattern detection
Rust Engine

Runs high-performance, memory-safe computation for large-scale microbiome datasets, long-read processing, indexing, and parallel execution.

  • High-speed sequence parsing
  • Long-read support
  • Parallel processing
  • Memory-safe computation
  • CLI and pocket app backend
  • Fast file format conversion
DUST Engine

Provides workflow automation, task orchestration, agent-based pipeline execution, and reproducible analysis coordination.

  • Automated workflow orchestration
  • Agent-based task execution
  • Pipeline scheduling
  • Tool coordination
  • Workflow logging
  • Reproducibility tracking
Jupyter Notebook Engine

Gives researchers an interactive notebook environment for exploration, reproducible analysis, custom scripts, visualization, and educational workflows.

  • Interactive microbiome analysis
  • Notebook-based reproducibility
  • Custom Python/R workflows
  • Research documentation
  • Visualization sandbox
  • Exportable reports
System Architecture Layers
L1
User Interface
Web Dashboard
CLI / Pocket Apps
Jupyter Notebook
L2
Workflow Layer
DUST Engine
MIU-INTEGRATE
MIU-Capsule
L3
Intelligence Layer
AI Engine
Consensus Engine
False Positive Shield
L4
Compute Layer
Rust Engine
Python Engine
Parallel Processing
L5
Data Layer
Amplicon
Shotgun WGS
Metatranscriptomics
Metaproteomics
Metametabolomics
Multi-Omics Results
EngineMain RoleStrengthUsed For
Python EngineScientific analysis
Flexible ecosystem
Data processing, statistics, ML, visualization
AI EngineDecision intelligence
Adaptive automation
Denoiser selection, false-positive detection, model scoring
Rust EngineHigh-performance compute
Speed and memory safety
Parsing, indexing, long-read processing, CLI backend
DUST EngineWorkflow orchestration
Automation and reproducibility
Pipeline control, task scheduling, agent execution
Jupyter EngineInteractive research
Researcher-friendly notebooks
Exploration, visualization, custom analysis
Jupyter Analysis Preview
import miu

project = miu.Project("gut_microbiome_study")

project.load_data("samples.fastq")
project.run_amplicon()
project.run_consensus()
project.ai.false_positive_shield()
project.export_report("MIU_Report.pdf")
Rust-Accelerated CLI Preview
miu run --input samples.fastq --omics amplicon \
       --engine rust --ai-consensus true --export report

MIU is not just a microbiome platform — it is a multi-engine computational framework built for scale, speed, and reproducibility.

Timeline

Project MIU Timeline

Year 1

Metagenomics Foundation

Phase 1A · Month 0–10
Amplicon + Shotgun Core
Phase 1B · Month 10–16
Long-Read + Real-Time
Year 2

Transcriptomics + Proteomics

Phase 2A · Month 16–22
Metatranscriptomics
Phase 2B · Month 22–28
Metaproteomics
Year 3

Metabolomics + Basic Integration

Phase 3A · Month 28–34
Metametabolomics
Phase 3B · Month 34–42
Multi-Omics Integration
Year 4

Production Release

Phase 4A · Month 42–45
Web Interface
Phase 4B · Month 45–48
Production Release
Prototype

Inside the MIU Workspace

A glimpse at the unified analysis dashboard.

miu.app / project / gut-cohort-2024
Live preview
Upload Dataset
FASTQ, mzML, mzXML, FASTA
Select Omics Type
Amplicon · WGS · Proteomics · …
Run Consensus Analysis
Multi-tool ensemble
View Reproducibility Score
Per-layer metrics
Export Report
PDF · HTML · JSON
Workflow Status
MIU-Capsule containers
Consensus Pipeline
Step 3 of 5 — MetaPhlAn462%
Host Removal: Completed
Kraken2 Classifier: Completed
MetaPhlAn4: Running
MOFA Integration: Queued
Run Results
Reproducible
Taxonomy F1 Score91%
Metaproteomics Reproducibility83%
Metametabolomics Reproducibility56%
False Positives Removed128
Engine Status
  • Python Engine
    Active
  • AI Engine
    Optimizing
  • Rust Engine
    Running
  • DUST Engine
    Orchestrating
  • Jupyter Engine
    Ready
Real-Time Pipeline Flow
Upload Dataset→
Python Preprocessing→
Rust Acceleration→
AI Consensus→
DUST Orchestration→
Jupyter Report→
Export Results
Pipeline progress4 / 7 stages