Microbiome Intelligence¶

Minting the next generation of microbiome analysis.

Open, SQL-native tools for microbiome data — from raw sequence and alignment to phylogeny, diversity, and mass spectrometry — built for the scale of modern multi-omic studies.

Explore duckdb-miint About Qiita

Projects¶

duckdb-miint Available now

A DuckDB extension that brings columnar analytics to microbiome research. Read and write FASTA/FASTQ, SAM/BAM, mzML, Newick and more; run alignment, denoising, chimera detection, classification, and UniFrac — all in SQL, all fast.
```
LOAD miint;
SELECT * FROM read_alignments('sample.bam')
WHERE mapq >= 30;
```
Documentation · Source
Qiita In active development

A scalable multi-omic platform for study management, processing, and analysis — amplicon, metagenomic, metatranscriptomic, metabolomic, and proteomic data. Designed for millions of samples and hundreds of terabytes, with duckdb-miint as its analytical engine.

Overview · Source
RYpe: Human DNA Filtration In active development

Remove human reads from a FASTQ without the reads leaving your computer. RYpe runs entirely in the browser: a k-mer screen against a human pangenome index (94 HPRC haplotypes, hg38, T2T-CHM13 v2.0, GENCODE v47 transcripts), then a minimap2 pass against T2T-CHM13 v2.0 over what survives. The index downloads once and stays in the browser.

Open RYpe · Source

Why MIINT¶

Columnar & fast

Microbiome data lives in big tables. MIINT meets it there — scanning and aggregating millions of records with DuckDB's vectorized engine instead of row-by-row scripting.
Verifiably correct

Every operation is test-driven and validated against reference tools. When a result matters, it is checked — not assumed.
Interoperable

Bridges the file formats and algorithms central to microbiome studies into DuckDB's ecosystem — BIOM, SAM/BAM, Newick, mzML, ENA/NCBI, and Parquet.
Open

Built in the open at the-miint, standing on a foundation of established, best-in-class bioinformatics tools.