---
title: Genomics FAQ
description: Frequently asked questions about genomics database tools on the platform.
tags:
  - genomics
  - faq
---

These answers cover how the [genomics](/docs/genomics) database tools differ and when to use each — Entrez versus Ensembl, KEGG versus Reactome — along with data freshness, non-human queries, interpreting ClinVar and OpenTargets results, downloading sequences, and chaining results into other platform tools.

## Genomics FAQ

## What is the difference between Entrez Search and Ensembl Gene Lookup?

Entrez Search queries any of 10 NCBI databases by keyword (gene, protein, nucleotide, etc.) and returns document summaries. It is broad and flexible. Ensembl Gene Lookup is specialized for gene coordinate retrieval: given a gene symbol or Ensembl ID, it returns precise genomic coordinates, biotype, assembly, and optionally transcript isoforms. Use Entrez for general search across databases; use Ensembl when you need coordinates and transcript details.

## When should I use KEGG versus Reactome for pathways?

KEGG is best for quick pathway keyword searches and metabolic network exploration. It covers pathways, genes, compounds, and reactions in a unified format. Reactome is best when you need manually curated, peer-reviewed pathway diagrams with detailed reaction mechanisms and disease annotations. Reactome's pathway browser provides richer visualization. For most initial pathway exploration, start with KEGG; for detailed mechanistic understanding, use Reactome.

## How current is the data returned by these tools?

Each database has its own update schedule. NCBI Entrez databases are updated daily. Ensembl has quarterly releases. KEGG is updated regularly (no fixed schedule). ClinVar is updated weekly with new variant submissions. OpenTargets has quarterly data releases. Reactome releases updates approximately quarterly. The tools always query the latest available data from each API.

## Can I search for non-human genes?

Yes. KEGG supports organism codes (`mmu` for mouse, `rno` for rat, `sce` for yeast, and many more). Ensembl supports species names (`mus_musculus`, `rattus_norvegicus`, `danio_rerio`, etc.). Entrez Search accepts organism filters via `[ORGN]` field tags (e.g., `BRCA1[GENE] AND mouse[ORGN]`). OpenTargets focuses on human disease associations only.

## What does "Variants of Uncertain Significance" (VUS) mean in ClinVar?

A VUS classification means there is insufficient evidence to determine whether the variant is disease-causing or benign. This is the most common classification in ClinVar. VUS variants should not be used for clinical decision-making without additional evidence. Over time, some VUS variants are reclassified as pathogenic or benign as more data becomes available.

## How are OpenTargets association scores calculated?

OpenTargets combines evidence from multiple data types into an overall score between 0 and 1. The four main evidence categories are: genetic association (GWAS, rare disease genetics), somatic mutation (cancer driver mutations), known drug (approved or clinical-stage drugs), and literature (text mining from publications). Each category is scored independently, and the overall score is an aggregation. A score above 0.7 indicates a well-validated target-disease association.

## Can I download full sequences?

Yes. Ask Ivy for the FASTA sequence of a protein or nucleotide record and give the UID or accession (for example "Get the FASTA sequence for protein accession NP_000537.3"); Ivy runs an Entrez fetch against the protein or nucleotide database. Results are returned as text content. Very large sequences (>50,000 characters) are truncated, but this covers most individual gene/protein records.

## How do I chain genomics results with other platform tools?

Common chains include: gene lookup followed by OpenTargets for target validation, then Target Passport for a comprehensive report. Or: Ensembl gene lookup followed by Entrez Link to find protein records, then UniProt fetch for annotation, then AlphaFold for structure prediction. Ivy handles chaining automatically when you describe a multi-step workflow.

## Read next