Unreported, unassessed, unresolved, and formally negative remain different states.
HOW TO READ THE EVIDENCE
The number, source, and time window travel with the claim.
This site uses a 61-person cohort snapshot with evidence reviewed through July 23, 2026. This page contains the few rules needed to interpret the rest of the site.
People, pedigrees, variants, source rows, observations, and model findings are counted separately.
Timing can define a testable question; it cannot prove mechanism or treatment effect.
COUNTS USED ON THIS SITE
Three compact denominator groups.
Use the labeled subset shown with a figure or statement. Do not read a detailed-case finding as a 61-person frequency.
01
People & evidence depth
How much source material supports the claim?
- 61
- active roster
- 51
- with mapped phenotype evidence
- 20
- record-backed clinical minimum
- 15
- deep longitudinal profiles
02
Molecular analysis
Which records can answer a sequence or variant-class question?
- 59
- exact molecular line visible
- 49
- sequence-defined
- 32
- LoF / truncating
- 18
- missense
- 3
- complex / mixed
- 5
- CNV / deletion
- 2
- protein-undefined splice
- 1
- pending
03
Separate contexts
Visible in the corpus, but not silently converted into independent recurrence.
- 5
- familial-cluster rows
- 17
- co-finding-control rows
- 21
- formal molecular reports
- 9
- direct molecular + clinical subset
Protein and domain analysis: sequence-defined records only.
Recurrence: related people and duplicate public rows are not independent events.
Clinical depth: 20 people meet the strict record-backed clinical minimum; 9 meet the narrower direct molecular-plus-clinical subset. The 21 formal molecular reports answer a different question.
Age: the current-age comparison includes 17 people age 18 or older. The earlier adult/transition source flag contains 16; the difference is one later age-confirmed participant, not a biological change.
Clinical matrix: 411 positive only + 55 positive with negative/unresolved + 12 negative only + 9 unresolved without positive + 428 unreported = 915 person-lane cells. Formal evidence is recorded within 102 of those cells; it is not an additional person count.
EVIDENCE LABELS
Six labels prevent unlike evidence from being blended.
Formal record
Original report, laboratory result, imaging, or signed clinical record.
Parent-reported formal content
Formal result relayed by a caregiver before the original record is available.
Parent-observed history
Direct observation with timing, context, recurrence, and recovery preserved.
Visual or media candidate
A visible feature in an image or video; mechanism remains unassigned.
Administrative or permission record
Consent, contact, source ownership, translation, or sharing status.
Inference or hypothesis
Interpretation or proposed mechanism—not an observed clinical fact.
Boundary marker: “Interpretive boundary” can appear beside a longitudinal summary. It is not a seventh evidence tier; it states what the available F, R, P, V, A, or X material cannot establish.
CHRONOLOGY
A clinical change is read in six parts.
The same sequence is used across development, seizures, movement, sleep, oral–GI, immune, cardiac/autonomic, medication, anesthesia, and daily function.
- T0Baseline
What was usual before the change?
- T1Context
What physiological or environmental window was present?
- T2Change
What exactly changed, for how long, and in which lane?
- T3Measurement
What record or test could answer the question?
- T4Intervention
What changed next, including no intervention?
- T5Recovery
Was return complete, partial, delayed, recurrent, or absent?
PUBLIC BOUNDARY
De-identified and aggregate by default.
Public-source records retain direct citations. Connected-cohort findings are shown at the broadest useful resolution. Names, exact private variants, dates, countries, family structures, and uniquely linkable combinations require documented permission and an appropriate sharing pathway.
Discuss a research or data request