Skip to content

Data sources & attribution

Yeliztli's own source code is released under the MIT license. To annotate your genome it draws on several public scientific datasets, each retained under its own license. This page summarises them; the repository NOTICE file holds the full, authoritative attribution text, and LICENSE covers Yeliztli's code.

For sizes and how each is obtained, see reference data.

Bundled / redistributed data

These are packaged with Yeliztli (or its release bundles), so their licenses permit redistribution.

Source Used for License Key citation
gnomAD v2.1.1 (GRCh37) Population allele frequencies, observed allele counts, and homozygous counts only CC0 1.0 Karczewski et al., Nature 581:434 (2020), doi:10.1038/s41586-020-2308-7
ClinGen gene–disease validity Context-only gene-validity guardrail CC0 1.0 (attribution requested) Strande et al., AJHG 100:895 (2017), doi:10.1016/j.ajhg.2017.04.015
AlphaMissense (hg19) Missense-pathogenicity context (complements REVEL) CC-BY-4.0 (per Zenodo record 10813168) Cheng et al., Science 381:eadg7492 (2023), doi:10.1126/science.adg7492
PharmGKB (LoE + DPWG) Cross-source pharmacogenomic evidence, context-only CC-BY-SA-4.0 (share-alike honored) Whirl-Carrillo et al., CPT 110:563 (2021), doi:10.1002/cpt.2350
FDA PGx labeling (via PharmGKB) Drug-label testing-level classification, context-only CC-BY-SA-4.0 (PharmGKB derivative; FDA text is public-domain)
GTEx v8 Tissue eQTL regulatory context (summary stats only) Open-access summary statistics GTEx Consortium, Science 369:1318 (2020), doi:10.1126/science.aaz1776
PGS Catalog (3 bundled scores) Polygenic-score weights — research-use only CC-BY-4.0 (per score) Lambert et al., Nat. Genet. 56:1989 (2024), doi:10.1038/s41588-024-01937-x

Only redistributable scores are bundled

Of the PGS Catalog, only the clearly redistributable CC-BY-4.0 scores are bundled — PGS000713 (type-2 diabetes), PGS005198 (BMI), and PGS000688 (LDL cholesterol). Non-commercial or author-restricted scores (such as the eBMD score, PGS000657) are not bundled and must be fetched by the user. Yeliztli deliberately excludes academic-license-restricted predictor columns (SpliceAI, CADD, REVEL, SIFT, PolyPhen) from the gnomAD bundle.

Downloaded from the provider

These are fetched from their original sources during setup/updates and used locally; they are not redistributed by Yeliztli.

Source Used for License
ClinVar (NCBI) Clinical variant classifications Public domain
dbNSFP In-silico predictions (REVEL, CADD, …) Academic / non-commercial
CPIC Pharmacogenomic allele & guideline data CC0-1.0
PharmVar Star-allele definitions Open
GWAS Catalog (EBI) Trait/disease associations Open
dbSNP (NCBI) rsID identity resolution Public domain
Mondo / HPO (Monarch) Disease & phenotype associations Open
UCSC Genome Browser hg19 + RefSeq (refGene) Optional local Genome Browser reference sequence and gene track UCSC Genome Browser data terms
Ensembl VEP Variant consequence predictions (the VEP bundle) Open (Ensembl)

Mind the licenses for your use

Some sources — notably dbNSFP (academic/non-commercial) and the share-alike PharmGKB/FDA-derived data — carry conditions. If you build on Yeliztli or its data, make sure your use complies with each source's terms.

Independence

"gnomAD", "Broad Institute", "ClinGen", "AlphaMissense", "Google DeepMind", "PharmGKB", "DPWG", "FDA", "GTEx", "PGS Catalog", "UCSC Genome Browser", "RefSeq", and other names appear here solely for source attribution. Yeliztli is an independent project and is not affiliated with, sponsored by, or endorsed by any of these organisations.