Evaluate & Confirm Results

The output of a Chromosomal Microarray (CMA) is a massive dataset containing thousands of data points. The analysis software (e.g., ChAS, BlueFuse) processes this data into a graphical visualization, typically a “Log2 Ratio Plot” and an “Allele Difference Plot” (for SNPs). The role of the clinical laboratory scientist is to filter this data, distinguish true biological signals from technical noise, and determine the clinical significance of any findings

Step 1: Quality Control (QC) Evaluation

Before looking at patient data, the run quality must be assessed. If the QC metrics fail, the data is unreliable (“Garbage In, Garbage Out”)

• Derivative Log Ratio Spread (DLRS): Measures the “noise” or “fuzziness” of the probe signals
  • Low DLRS (<0.2): Good quality. The probes are tight; small deletions are visible
  • High DLRS (>0.3): Poor quality. The data is “noisy.” The scatter is so wide that it might hide a true deletion or create false positives
• SNP QC: Measures the ability of the array to distinguish between Genotype A and Genotype B
• Gender Match: Does the array gender (XX/XY) match the patient requisition? A mismatch indicates a sample swap (Mix-up)

Step 2: Identifying Copy Number Variants (CNVs)

The analyst scrolls through the genome (Chromosome 1 through Y) looking for deviations from the baseline

• Log2 Ratio Plot
  • Baseline (0): Normal (2 copies). \(\log_2(2/2) = 0\)
  • Gain (+0.58): One extra copy (3 copies total). \(\log_2(3/2) = 0.58\). The dots shift UP
  • Loss (-1.0): One missing copy (1 copy total). \(\log_2(1/2) = -1.0\). The dots shift DOWN
  • Homozygous Loss (Null): Zero copies. \(\log_2(0/2) = \text{Undefined (or extremely low)}\). The dots drop to the bottom floor
• B-Allele Frequency (SNP Plot): Used to confirm the Copy Number call
  • Normal: 3 tracks (AA at 1.0, AB at 0.5, BB at 0)
  • Deletion: The middle “AB” track disappears (you cannot have Heterozygosity if you only have 1 copy). Tracks compress to only AA (1.0) and BB (0)
  • Duplication: The heterozygous track splits into two: AAB (0.66) and ABB (0.33)

Step 3: Clinical Interpretation (The Filter)

Once a CNV is found (e.g., “300kb deletion on 16p11.2”), is it pathogenic?

• Size: Generally, larger CNVs (>500kb) are more likely to be pathogenic. Small CNVs (<50kb) are often benign noise
• Gene Content
  • Does the deletion encompass “OMIM Morbid” genes (genes known to cause disease)?
  • Does it contain haploinsufficient genes (genes that break if you lose just one copy)?
• Databases
  • ClinGen / DECIPHER: Databases of pathogenic variants. If the deletion matches a known syndrome (e.g., DiGeorge, Prader-Willi), it is likely Pathogenic
  • DGV (Database of Genomic Variants): A database of benign variations found in healthy people. If the deletion is found in 5% of healthy adults, it is Benign
• Classification
  • Pathogenic: Causes the disease
  • Likely Pathogenic: 90% chance it causes disease
  • Variant of Uncertain Significance (VUS): Insufficient evidence
  • Likely Benign / Benign: Does not cause disease

Step 4: Confirmation (The Second Method)

In clinical genetics, a significant finding (especially a new diagnosis) must often be confirmed by an independent method to rule out sample mix-up or technical artifact

• FISH Confirmation
  • If the CNV is large enough (>100kb), FISH is the preferred confirmation method
  • Why? It confirms the presence of the deletion/gain visually and proves it is in the patient’s cells. It also allows for Parental Studies: (testing Mom and Dad) to see if the defect was inherited or de novo
• qPCR / MLPA
  • Used for smaller CNVs that are below the resolution of FISH probes
• Karyotype
  • If the array shows a Gain, a karyotype might be ordered to see where the extra material is (e.g., is it a ring chromosome? An unbalanced translocation?)

Step 5: Reporting

The final report must include:

• ISCN Nomenclature: arr[GRCh37] 22q11.21(18,912,231-21,465,672)x1
  • Describes the exact genomic coordinates (start-stop) and the copy number (x1)
• Interpretation: A plain English explanation (“This result is consistent with 22q11.2 Deletion Syndrome…”)
• Recommendation: “Parental testing is recommended to determine recurrence risk.”