How to make a pedigree chart in 5 steps

A clinical-grade, step-by-step construction: identify the proband, draw the parents, add siblings, extend to three generations, annotate affected status. With ASCII diagrams at each step and the common pitfalls that catch novices out.

| 12 min read

Short version. A clinical pedigree is built in five disciplined steps. (1) Identify the proband and mark with an arrow. (2) Draw the parents on a horizontal mating line above the proband. (3) Add siblings on a horizontal sibship line at the same generation. (4) Extend vertically to grandparents so the chart covers three generations. (5) Annotate affected status, ages, and test results. The result should be readable by another clinician without any verbal explanation — if it isn't, something has been missed.

The steps below assume NSGC 2022 notation (Bennett et al., Journal of Genetic Counseling). If you are new to the symbol vocabulary, read our pedigree symbols reference first; the rest of this page takes the vocabulary for granted.

Step 1: Identify the proband

The proband is the first affected individual who brought the family to clinical attention. Everything that follows is structured around that individual. Mark the proband with an arrow pointing to the lower-left of their shape.

  Start here:
         ┌───┐
         │   │   ← proband placeholder (sex known? fill in shape)
         └───┘
          ↙

If a different family member is the consultand — the person who asked for genetic counselling, but who is not the first-affected individual — they are marked separately with the letter P. Many family histories, particularly in cancer genetics, come from an unaffected daughter asking about her mother; the mother is the proband, the daughter is the consultand.

Pitfalls at this step.

  • Forgetting the arrow. Without it, a chart cannot be clinically interpreted.
  • Confusing proband with consultand. They are frequently the same person but not always.
  • Picking the wrong proband in large families. By convention, the proband is the index case — the person whose diagnosis prompted the referral — not the most severely affected family member.

Step 2: Add the parents on a horizontal mating line

Draw the proband's parents on a horizontal line above the proband's generation. A horizontal mating line connects the two parents. A vertical line descends from the mating line to the proband's sibship line.

      Father    Mother           (generation II)
       ┌───┐─────○
       │   │    │
       └───┘    ╰
              │
              │ (descent line)
              │
              ●  ← PROBAND        (generation III)
             (affected)

Place the male partner on the left, the female on the right, by the usual convention — though this is a convention for consistency, not a clinical requirement. Draw each parent's shape with the correct sex symbol (square for male, circle for female, diamond for unknown / non-disclosed sex) and record affected status, age, and any relevant family history below the shape.

Pitfalls at this step.

  • Drawing the parents as siblings. Siblings share a sibship line and descend from a mating line above; partners are on a mating line themselves, without an ancestry line descending onto them (their own ancestry is drawn above when you extend to three generations).
  • Missing parent ancestry. When both parents have relevant family history, both need their own ancestry extended upward (both sets of grandparents).
  • Recording "unknown" parents as "unaffected". Unknown status should be noted explicitly — a question mark inside the shape, or an annotation — not silently conflated with unaffected.

Step 3: Add siblings on a horizontal sibship line

The proband's siblings branch from the same sibship line as the proband. Draw siblings in birth order, oldest on the left, each connected by a short vertical line descending from the horizontal sibship line.

              Father      Mother
               ┌───┐───────○
               │   │       │
               └───┘       ╰
                          │
               ┌──────────┼──────────┐
            (oldest)                (youngest)
               │          │          │
              ┌───┐      ●         ┌───┐
              │   │      ← PROBAND  │   │
              └───┘                  └───┘
              III-1    III-2       III-3
             (unaff.) (affected)  (unaff.)

Include all live-born siblings and, where clinically relevant, reproductive losses (spontaneous abortions, stillbirths, terminations) — these are drawn using the triangle and annotated shapes described in the symbols reference. Reproductive losses are often clinically relevant: a family with multiple miscarriages may suggest a balanced translocation in one parent, for example.

Pitfalls at this step.

  • Drawing only the live-born children. Reproductive losses carry clinical information and should be recorded if known.
  • Mixing the order. Birth order matters for interpretation and for later reference.
  • Placing half-siblings on the same sibship line as full siblings. Half-siblings from a different partner of one parent require a second mating line from that parent to the other partner, with a separate sibship line below.

Step 4: Extend upward to three generations

A clinical pedigree covers three generations at minimum: the proband's generation (III), the parents' generation (II), and the grandparents' generation (I). Extend each parent upward to their own parents — i.e. the proband's grandparents — plus any aunts and uncles on each side.

   I-1    I-2              I-3     I-4         Generation I
    ┌───┐──○                ┌───┐───○           (grandparents)
    │   │                   │   │
    └───┘                   └───┘
      │                        │
  ┌───┴───┐               ┌────┴────┐
 II-1   II-2            II-3      II-4          Generation II
  ■      ○                □         ●           (parents and
 uncle  mother           father    aunt          their siblings)
        (proband's        (proband's
         mother)           father)
              │                │
              │                │
              └────────┬───────┘
                       │
              ┌────────┼────────┐
            III-1     III-2    III-3            Generation III
             ○         ■        □                (proband and
             sister  PROBAND   brother            siblings)

Collect detailed information on at least the proband's first-degree relatives (parents, siblings, children) and second-degree relatives (grandparents, aunts, uncles, half-siblings). Third-degree relatives (first cousins, great-grandparents) are useful for conditions with variable penetrance or when tracing a founder variant. For hereditary cancer assessment, at least three generations on both sides of the family are clinical practice — cancer clusters on either side matter.

At this step, the overall chart topology is set. It is worth looking at the chart and asking: does it display the family in a way that makes the clinical question easy to read? Branches can sometimes be rearranged (within the constraints of correct ancestry) to improve clarity.

Step 5: Annotate affected status and ages

Fill each shape according to affected status for the condition(s) being tracked. Annotate ages and, for affected individuals, the phenotype and age of onset.

      John                 Mary
      III-1                III-2
      ┌───┐                ●
      │   │                ← PROBAND
      └───┘                (BrCa, dx 42)
      42y                  40y
      unaff.               BRCA1 c.5266dupC (VCF: pathogenic)

  Key: ■/● = affected (breast or ovarian cancer);
       ┌───┐ = unaffected; ↙ = proband;
       — = biological relationship; d. = deceased.

A well-annotated pedigree, per line below each shape:

  • Identifier: generation-position (e.g. III-2)
  • Name or initials
  • Age: current age for living, "d. [age]" for deceased
  • Affected status: condition and age of onset, e.g. "BrCa, dx 42"
  • Test results: gene, variant in HGVS nomenclature, classification (e.g. "BRCA1 c.5266dupC, pathogenic"). For unaffected tested individuals, record the predictive test result (e.g. "BRCA1 c.5266dupC, not detected").

Always add a legend on the chart defining what each fill pattern means. Where multiple conditions are tracked, use quadrant fills per the symbol reference and document each quadrant's meaning in the legend.

Pitfalls at this step.

  • Omitting ages. Age-dependent penetrance models cannot function without ages.
  • Confusing "unaffected" with "untested". An unaffected relative who has tested negative for a familial variant is different clinically from an unaffected relative who has not been tested. Annotate accordingly.
  • Using inconsistent fill patterns across conditions. If colon cancer is upper-left in one person, it must be upper-left everywhere on the chart.
  • Free-text phenotype labels. Use ICD-10 or OMIM codes where possible so the chart is machine-readable and interoperable.

Common pitfalls across all five steps

Pitfall Why it matters Fix
No proband arrowChart can't be clinically interpretedAlways draw the arrow as step 1
Missing agesBreaks age-dependent penetrance modelsRecord current age or age at death for every individual
Unclear affected statusMisclassifies risk in risk modelsDistinguish "unaffected" from "unknown status" explicitly
Too few generationsInheritance pattern cannot be establishedCover three generations minimum
Mixing half-siblings as full siblingsMisrepresents shared ancestryUse a separate mating line and sibship line per partnership
No legendFill patterns become ambiguousAlways add a legend, even for single-condition pedigrees
Free-text phenotypesNot machine-readable, risks ambiguityUse ICD-10 or OMIM codes

How Evagene supports this

Evagene's gesture drawing mode automates the layout that the five steps produce. A clinician marks the proband — the arrow appears automatically and the proband is placed near the centre of the canvas. Gestures add parents above, siblings to the sides, and children below, with the mating and sibship lines drawn automatically and correctly. The three-generation scaffold expands as more family members are added, and the layout rebalances to keep the chart legible without redrawing.

Symbol fills, quadrant logic for multiple conditions, consanguinity detection, and reproductive-loss shapes are applied automatically based on what the clinician records in the individual's annotation pane. The legend is generated from the condition list on the chart, so it is always consistent with the fills.

Because the chart is data-backed, it is exportable as GEDCOM (for archival and cross-institution transfer), PDF (for the patient record), SVG (for embedding in reports or patient portals), and JSON (for programmatic access via our REST API). A chart drawn in Evagene in the first visit is never re-drawn from scratch on the second — it is loaded, extended, and re-exported.

Frequently asked questions

What is the first step in drawing a pedigree?

Identify the proband and mark with an arrow pointing to the lower-left of the shape. Place the proband centrally so the chart can expand around them.

How many generations should a clinical pedigree include?

Three minimum: proband, parents, grandparents. Add children if any. Extend to four where information is available and clinically relevant.

What is the difference between a mating line and a sibship line?

A mating line connects partners. A sibship line is the line from which siblings descend. A vertical line joins the two.

How do you record ages on a pedigree?

Below the shape: current age, or "d." with age at death. Ages drive age-dependent penetrance calculations.

Should you always draw affected individuals with filled shapes?

Yes for single-condition pedigrees. For multiple conditions, use quadrant fills with a legend. Distinguish "unaffected" from "unknown" with a question mark.

What is the most common mistake?

Missing the proband arrow, followed by conflating mating and sibship lines, followed by omitting ages.

Do I need software to draw a clinically valid pedigree?

No — standard notation is software-agnostic. Software helps with version control, re-export, risk-model integration, and automatic standardisation.

Related reading

Evaluate Evagene for your service

Join the Alpha waiting list. No credit card, no enterprise sales cycle — free access during Alpha for clinicians and research teams.

Join the Alpha Waiting List