Skip to main content Link Menu Expand (external link) Document Search Copy Copied
Biology Skills

Biological identity and plot labels

Summary

Every data point in a biology plot represents a specific entity (e.g. a gene, transcript, variant, sample). Labels and annotations must preserve the full identity. For genetic variants, use exact HGVS notation tied to a reference sequence. Do not replace identifiers with shorthand or symbols that lack context. If the underlying analysis uses a specific transcript or isoform (e.g. MANE Select), mention it. Omitting these details can mislead or prevent reproducibility.

Core rules

  • Exact naming: Label points/lines with full identifiers (e.g. NM_000546.6:c.524G>A (p.Arg175His)), not ambiguous short forms. Preserve HGVS prefixes (g., c., p.) appropriately.
  • Reference context: Include versioned accession and genome build when relevant. A gene symbol alone is not a substitute for a sequence reference.
  • Consistent symbols: Use approved gene symbols only to supplement, not replace, precise labels. Do not embed symbols in HGVS strings (keep them separate).
  • Metadata annotation: If plotting summary statistics, annotate axes or legends with sample source (e.g. tissue, donor ID) so the biological origin is clear.
  • Color/shape legends: Legends encoding categories must clearly map to the exact categorical values (e.g. tissue type names, not index numbers).

Required context

  • Full accession/version of sequences (genome, transcript, protein) referenced by data.
  • Source of each data point (sample ID, cell line, patient, etc.).
  • If labels derive from pipeline (e.g. gene from differential analysis), know which reference was used (Ensembl or RefSeq ID, and version).

AI behaviour

  • Preserve identifiers: Do not truncate or generalize. If asked “What mutation is this?”, answer with complete HGVS description and reference accession (or state it is missing).
  • Check consistency: When summarizing groups (e.g. “genes upregulated”), ensure each name is uniquely defined (disambiguate synonyms).
  • Report missing context: If reference sequence or build is not known, note that labels may be incomplete.

Common failure modes

  • Shorthand labels: Plot uses “p.Arg132Leu” without transcript/coordinfo. Reader cannot tell which variant.
  • Missing reference: Using chromosome coordinates (chr1:123456) without specifying assembly (GRCh37 vs GRCh38) makes it non-reproducible.
  • Inconsistent identifiers: Mixing gene symbols and protein IDs in one figure, causing confusion about units.
  • Color-coded without legend: Categories (e.g. cell types) shown only by colours; without a key, meaning is lost or misinterpreted.

Authoritative standards

  • HGVS Nomenclature: Adopt the HGVS rules for variant descriptions (e.g. include “c.” or “p.” prefix, no spaces).
  • MANE/Reference IDs: Use MANE Select or RefSeq accession for transcripts; for proteins use UniProt or RefSeq when needed.
  • Gene nomenclature: Follow HUGO Gene Nomenclature (HGNC) for gene symbols if included, but never as sole label.

Examples

Example: Variant point plot

  • Data: Variants on a gene vs. effect size.
  • Check: Label each point with full c.HGVS and p.HGVS or tooltip. Axis label “Amino acid change (HGVS p.)” plus caption “ref NM_XXX”.
  • Good outcome: X-axis tick labels are “p.(Arg175His)”, legend note “transcript NM_000546.6 (MANE)” (sources: HGVS standard).

Example: Sample category colouring

  • Data: Bars colored by tissue type.
  • Check: Legend entries are full tissue names (“Lung”, “Liver”), not codes; axis label “Sample ID” has prefix if needed (“Sample123_A”).
  • Good outcome: Legend says “Cell line origin: Lung, Liver, …”; each bar labeled with unique sample barcodes.

Sources

  • HGVS Nomenclature recommendations.
  • Biology Skills: variant nomenclature guidance (genomics).