Understanding Your Results
How to interpret Omics807 results - VCF filters, quality scores, Omics807 insights, and visualization
Understanding Your Results
Learn how to interpret Omics807's analysis results, quality metrics, and Omics807 insights.
Results Dashboard Overview
After analysis completes, Omics807 presents an interactive dashboard with:
- Summary Statistics - High-level metrics
- Variant Table - Detailed variant list with enrichment data
- Enrichment Data - Population genetics, protein structures, drug targets, pathways
- Omics807 Interpretations - Comprehensive clinical insights for top variants
- Visualizations - Chromosome distribution and quality charts
Summary Statistics
Key Metrics Explained
Total Variants Detected - All variants found in VCF file - Includes PASS, GERMLINE, and RefCall - Typical range: 1,000-50,000+ for WGS
High-Quality Variants (PASS) - Variants passing all quality filters - Most likely to be true somatic mutations - Focus analysis on these variants
Chromosomes Affected - Number of chromosomes with variants - Expected: Most or all chromosomes for WGS - Limited for targeted/exome sequencing
Average Quality Score - Mean QUAL score across variants - Higher = more confident calls - Good threshold: QUAL > 30
VCF FILTER Status
DeepSomatic assigns FILTER values to classify variants:
PASS - Somatic Variant ✅
Meaning: High-confidence somatic mutation
Characteristics: - Present in tumor, absent (or low VAF) in normal - Passes all quality thresholds - Most likely cancer-specific
Example:
chr7 140753336 . A T 60.0 PASS DP=120;VAF=0.35
^^^^
High confidence somatic
Action: Prioritize for downstream analysis
GERMLINE - Inherited Variant 🧬
Meaning: Likely germline (inherited), not somatic
Characteristics: - Present in both tumor and normal - Similar VAF in both samples (~50%) - Part of patient's genome, not cancer-specific
Example:
chr1 12345 . G A 45.0 GERMLINE DP=80;VAF=0.48
^^^^^^^^
Similar VAF in normal/tumor
Action: Filter out for cancer analysis (unless cancer predisposition genes)
RefCall - Reference Call 📍
Meaning: No variant detected, matches reference
Characteristics: - Included for gVCF completeness - Confirms coverage at position - Can be safely ignored for variant analysis
Example:
chr2 98765 . C . . RefCall DP=60
^^^^^^^
No variant here
Action: Ignore for variant analysis
LowQual - Low Quality ⚠️
Meaning: Variant detected but below quality threshold
Characteristics: - QUAL score too low - Insufficient evidence - May be sequencing error
Action: Exclude from high-confidence analysis
Quality Scores Explained
QUAL - Variant Quality
Definition: Phred-scaled probability that variant exists
Formula: QUAL = -10 × log₁₀(P_error)
Interpretation:
QUAL = 10 → 90% confidence (1 in 10 chance of error)
QUAL = 20 → 99% confidence (1 in 100)
QUAL = 30 → 99.9% confidence (1 in 1,000)
QUAL = 60 → 99.9999% confidence
Thresholds: - Minimum: QUAL > 10 - Good: QUAL > 30 - Excellent: QUAL > 60
GQ - Genotype Quality
Definition: Confidence in the specific genotype call
Genotype Notation:
- 0/0 = Homozygous reference (normal)
- 0/1 = Heterozygous (one variant allele)
- 1/1 = Homozygous alternate (both alleles variant)
Interpretation:
GQ = 20 → 99% confident in genotype
GQ = 40 → 99.99% confident
GQ = 60 → 99.9999% confident
Use: Filter variants with GQ < 20 for uncertain genotypes
DP - Read Depth
Definition: Total number of reads covering the position
Importance: - Higher depth = more reliable calls - Low depth = potential false positive/negative
Thresholds:
DP < 10 → Too low, unreliable
DP = 20-30 → Acceptable for WGS
DP > 50 → High confidence
DP > 100 → Very high confidence (WES)
Warning: Extremely high DP (>500) may indicate: - Alignment artifacts - Repetitive regions - Copy number amplifications
VAF - Variant Allele Frequency
Definition: Percentage of reads supporting the variant
Formula: VAF = Variant Reads / Total Reads
Interpretation for Somatic Variants:
VAF = 100% → Homozygous, all tumor cells
VAF = 50% → Heterozygous (or germline)
VAF = 30% → 60% tumor purity, heterozygous
VAF = 10% → Subclonal, minor population
VAF < 5% → Very rare, low confidence
Factors Affecting VAF: - Tumor purity (normal cell contamination) - Copy number alterations - Subclonal populations - Sequencing depth
Example:
Tumor: DP=100, Variant reads=35 → VAF = 35%
Normal: DP=80, Variant reads=0 → VAF = 0%
→ Somatic variant, ~70% tumor purity
Variant Table Columns
Omics807 displays variants in a comprehensive table:
| Column | Description | Example |
|---|---|---|
| Chromosome | Genomic location | chr7 |
| Position | Coordinate (1-based) | 140753336 |
| Ref | Reference allele | A |
| Alt | Alternate allele | T |
| Type | Variant type | SNV |
| QUAL | Quality score | 60.0 |
| FILTER | Pass/filter status | PASS |
| DP | Read depth | 120 |
| VAF | Allele frequency | 0.35 (35%) |
| GQ | Genotype quality | 55 |
Sorting and Filtering
Recommended Filters: 1. FILTER = PASS only 2. QUAL > 30 3. DP > 20 4. VAF > 0.05 (5%)
Sort Priority: 1. By QUAL (highest first) - most confident 2. By VAF (highest first) - clonal variants 3. By chromosome/position - genomic order
Omics807 Insights Interpretation
Omics807 uses advanced analysis to provide comprehensive clinical context for top variants, integrating population genetics, protein structure analysis, drug targeting, pathway analysis, and clinical evidence.
Analysis Components
1. Clinical Significance - Pathogenicity assessment - Known cancer associations - Functional impact predictions
Example:
"This BRAF V600E mutation is a well-established
oncogenic driver in melanoma and colorectal cancer,
associated with constitutive MAPK pathway activation."
2. Associated Cancer Types - Primary cancers where variant is common - Frequency in cancer databases - Prognostic implications
Example:
"Most frequent in melanoma (50%), thyroid cancer (45%),
and colorectal cancer (10%). Generally associated with
poor prognosis but targetable with BRAF inhibitors."
3. Treatment Implications - FDA-approved therapies - Clinical trial options - Drug resistance markers
Example:
"Targetable with BRAF inhibitors (vemurafenib, dabrafenib)
often combined with MEK inhibitors. Consider resistance
testing for NRAS, MEK, and PTEN mutations."
4. Recommended Actions - Confirmatory testing - Additional assays - Clinical guidelines
Example:
"Confirm with orthogonal method (Sanger sequencing or ddPCR).
Consider comprehensive genomic profiling to identify co-mutations.
Refer to NCCN guidelines for treatment selection."
Limitations of Omics807 Insights
⚠️ Important Caveats: - Omics807 analysis is supplementary, not diagnostic - Always validate with clinical databases (COSMIC, ClinVar) - Consult with oncologists/genetic counselors - AI may not reflect latest research (knowledge cutoff)
Chromosome Distribution Chart
Reading the Visualization
The chromosome distribution chart shows:
X-axis: Chromosomes (1-22, X, Y) Y-axis: Variant count per chromosome
Typical Patterns:
Normal Distribution:
Variants
| ▄ ▅
|▅ █ ▆ █ ▆
|█▅█▆█▇█▆█▅▆
└────────────
1 2 3...22 X
- More variants on larger chromosomes
- Relatively even distribution
Focal Amplification:
Variants
| ██
| ▄ ▅ ██
|▅ █ ▆ █▆██▅▆
└────────────
^^
chr17 enrichment
- Spike on specific chromosome
- May indicate copy number gain
- Could be chr17 (HER2) amplification
Chromothripsis:
Variants
| ████
| ▄ ████ ▅
|▅ █▆████▆█▅▆
└────────────
^^^^
Massive rearrangement
- Extreme variant clustering
- Suggests catastrophic chromosome shattering
Clinical Interpretation
Chromosome 17 enrichment: - Check for TP53 mutations - Consider HER2 amplification (breast cancer)
Chromosome 13/14 enrichment: - Possible RB1 pathway alterations - Common in retinoblastoma, osteosarcoma
Sex chromosome variants: - X chromosome: Check BRCA1, AR - Y chromosome loss: Common in aging, some cancers
Quality Score Histogram
Understanding the Distribution
Ideal Distribution:
Count
| ▄▄▄█████
| ▄▄███████████
| ▄▄██████████████▄
| ▄▄██████████████████▄▄
└─────────────────────────
0 10 20 30 40 50 60
QUAL score
Most variants clustered at high QUAL (>40) ✅
Concerning Distribution:
Count
|██
|███▄
|█████▄
|███████▄▄▄▄▄▄
└─────────────────────────
0 10 20 30 40 50 60
Too many low-quality variants (<20) ⚠️
Action: May indicate: - Low coverage sample - Poor sequencing quality - Degraded DNA (FFPE)
Exporting and Sharing Results
Download Options
VCF File - Standard format for variant data - Compatible with all genomic tools - Can be uploaded to genome browsers
CSV Export - Spreadsheet-friendly format - Easy filtering in Excel - Good for non-bioinformaticians
PDF Report (if available) - Summary for clinical records - Shareable with healthcare providers - Includes visualizations
Next Steps After Analysis
1. Validation - Confirm key variants with Sanger sequencing - Use orthogonal platform (e.g., ddPCR for VAF)
2. Annotation - Annotate with VEP, ANNOVAR, or SnpEff - Add functional predictions (SIFT, PolyPhen) - Cross-reference with COSMIC, ClinVar
3. Clinical Correlation - Compare to patient's cancer type - Check for actionable mutations - Assess for clinical trial eligibility
4. Further Analysis - Copy number analysis (CNVkit, FACETS) - Mutational signature analysis - Pathway enrichment analysis
Troubleshooting Results
Too Few Variants
Expected: 1,000-10,000 for WES, 10,000-100,000 for WGS
If too low: - Check coverage (DP) - may be insufficient - Verify tumor purity - low purity = fewer calls - Confirm correct model used - Review FILTER column - variants may be filtered
Too Many Variants
If suspiciously high (>100,000):
Possible causes: - Wrong reference genome (GRCh37 vs GRCh38) - Contamination - Tumor-only without PoN filtering - Sequencing artifacts
No PASS Variants
All variants filtered: - Insufficient sequencing quality - Very low tumor purity - Germline contamination - Wrong model selection (e.g., WGS model on WES data)
Inconsistent VAF
Variant with VAF > 50% but marked somatic: - May indicate copy number gain - Could be LOH (loss of heterozygosity) - Verify tumor purity estimate
Learn More
- Model Guide - Choose the right model
- Genomics 101 - Understand the basics
- Glossary - Look up terminology
- FAQ - Common questions