Medicine
Genomics of Rare Diseases in Clinical Diagnostics
Quick fact
For rare diseases, whole-exome sequencing identifies a causative mutation in only about 25–30% of patients, meaning most patients still end their diagnostic odyssey without a molecular explanation.
Why this is interesting
When a child is born with a mysterious syndrome, it can take years—and dozens of wrong guesses—before a diagnosis is found. But sometimes, the answer is hiding in a single letter of the human genome.
Read the full explanation
Understanding Genomics of Rare Diseases in Clinical Diagnostics
Imagine the human genome as a vast instruction manual written in over 3 billion letters. Rare diseases are often caused by small typos in this manual. Clinical genomics uses powerful sequencing methods to read an individual's DNA. The most common approach is exome sequencing, which focuses only on the ~1% of the genome that codes for proteins. This is strategic because most known disease-causing mutations lie in these coding regions. Doctors sequence the exome of a patient and often their parents too. By comparing the patient's genetic code to the parents', they can spot novel mutations. When they find a mutation in a gene already known to cause the patient's symptoms, they can make a genetic diagnosis. This is a stark contrast to the traditional, often agonizing, diagnostic odyssey of clinical exams and metabolic tests.
A deeper explanation
The mechanism underlying clinical genomics for rare disease is built on the principle that these conditions are 'Mendelian'—caused by mutations in a single gene with high penetrance. By sequencing exomes, scientists search for rare, damaging variants that segregate with the disease within the family. Variant interpretation follows structured guidelines (like ACMG/AMP) that combine multiple lines of evidence: allele frequency in the general population, predicted effect on the protein (e.g., truncation), the variant's presence in other affected individuals, and functional assays. Data-sharing across global databases is critical to accumulate these rare variants and link them to phenotypes. However, the diagnostic yield is limited by several factors: the mutation may be in non-coding regions, could be a structural variant (deletion/duplication) missed by exome sequencing, or the disorder may be caused by a combination of variants (oligogenic). Additionally, variants may be found of 'uncertain significance' where clinical action is unclear. The consequence is that while genomics can end the diagnostic odyssey for some, it introduces a translational bottleneck: turning a genetic variant into a clinically actionable diagnosis requires robust bioinformatics, clinical correlation, and ethical handling of incidental findings (like cancer risk variants).