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:

  1. Summary Statistics - High-level metrics
  2. Variant Table - Detailed variant list with enrichment data
  3. Enrichment Data - Population genetics, protein structures, drug targets, pathways
  4. Omics807 Interpretations - Comprehensive clinical insights for top variants
  5. 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

References