Documents

1. Sequencing & Variant Calling

Tumor-only somatic variant analysis using targeted panel sequencing, aligned to GRCh37 (hg19) reference genome.

Panel Types:

2. Variant Filtering (VCF Processing)

• • Genotype (GT) extraction and DP > 100 filtering

• • Multi-allelic variants split into separate records

• • VAF (Variant Allele Frequency) calculation

• • VAF ≥ 0.05 filtering threshold

3. Variant Annotation (VEP)

Annotation using Ensembl VEP (Variant Effect Predictor) with the following criteria:

• Transcript Selection

  Canonical transcript only (Canonical = Yes)

• Population Frequency Filtering

  • gnomAD_exon_EAS_AF < 0.001
  • gnomAD_genome_EAS_AF < 0.001
  • MAX_AF < 0.01

• Splice Site Prediction

  SpliceAI cutoff ≥ 0.5 included

• Clinical Significance

  ClinVar: Exclude Benign variants

• Biotype

  Protein coding genes only

• COSMIC Filtering

  Remove variants with COSMIC Germline = Yes and Somatic = No

• Consequence Selection

  missense_variant

  stop_gained

  stop_lost

  start_lost

  frameshift_variant

  inframe_deletion

  inframe_insertion

  splice_acceptor_variant

  splice_donor_variant

4. Statistical Analysis

• • Kaplan-Meier survival analysis (OS and PFS)

• • Log-rank test for group comparison

5. Visualization & Reporting

• • OncoPlot generation for mutation landscape

• • Lollipop plots for protein-level mutations

• • Kaplan-Meier survival curves

• • Disco Plot for individual sample genomic overview (SNV/Indel, CNV, Fusion on circular chromosome layout)

• • Interactive data exploration interface

Sample Data

Survival Data

Status: 0 = censored (alive or no progression), 1 = event (death or progression)

Mutation Data & Annotation

Each variant is annotated with multiple prediction tools and databases:

• Basic Information

  Gene Symbol, Chromosome, Position, Reference/Alternate Allele, Variant Type (SNP, INS, DEL)

• Consequence (Variant Effect)

  Predicted effect of variant on gene/transcript:
  • missense_variant: Amino acid changed to different amino acid
  • stop_gained: Premature stop codon created (nonsense mutation)
  • stop_lost: Stop codon removed, protein extension
  • start_lost: Start codon removed
  • frameshift_variant: Insertion/deletion causing reading frame shift
  • inframe_deletion: Deletion preserving reading frame
  • inframe_insertion: Insertion preserving reading frame
  • splice_acceptor_variant: Variant in splice acceptor site (3' end of intron)
  • splice_donor_variant: Variant in splice donor site (5' end of intron)

• Impact (Severity)

  Predicted severity of variant effect on protein function:
  • HIGH: Likely loss of function (frameshift, stop_gained, splice variants)
  • MODERATE: Possible functional change (missense, inframe indels)
  • LOW: Unlikely to affect function significantly (synonymous)
  • MODIFIER: Non-coding or intergenic variants

• HGVS Notation

  HGVSc (coding DNA change, e.g., c.1234A>G), HGVSp (protein change, e.g., p.Val412Met) - Standardized nomenclature for describing variants

• SIFT (Sorting Intolerant From Tolerant)

  Predicts whether an amino acid substitution affects protein function based on sequence homology.
  • Tolerated: Score ≥ 0.05 (high confidence)
  • Tolerated Low Confidence: Score ≥ 0.05 (low confidence)
  • Deleterious: Score < 0.05 (high confidence)
  • Deleterious Low Confidence: Score < 0.05 (low confidence)
  • Unknown: No prediction available

• PolyPhen-2 (Polymorphism Phenotyping v2)

  Predicts the impact of amino acid substitutions on protein structure and function.
  • Benign: Score < 0.15 (likely neutral)
  • Possibly Damaging: Score 0.15-0.85 (potential functional impact)
  • Probably Damaging: Score > 0.85 (likely deleterious)
  • Unknown: No prediction available

• REVEL (Rare Exome Variant Ensemble Learner)

  Ensemble method combining multiple tools to predict pathogenicity of missense variants.
  • Benign: Score ≤ 0.644 (likely neutral)
  • Pathogenic: Score > 0.644 (likely disease-causing)
  • Unknown: No prediction available

• SpliceAI

  Deep learning-based tool that predicts splicing alterations by evaluating delta scores for donor/acceptor gain and loss. Variants with any delta score ≥ 0.5 are marked as PASS, indicating a high likelihood of affecting splicing.

• dbSNP (Database of Single Nucleotide Polymorphisms)

  NCBI database cataloging known genetic variations (rs IDs). Presence indicates previously reported variant in general population

• COSMIC (Catalogue Of Somatic Mutations In Cancer)

  Comprehensive database of somatic mutations in cancer. COSMIC ID indicates variant has been observed in cancer samples

• ClinVar

  NCBI database of clinically significant variants. Classifications: Pathogenic, Likely Pathogenic, Uncertain Significance, Likely Benign, Benign

Log-Rank Test

Statistical test to compare survival distributions between two or more groups. A p-value < 0.05 indicates statistically significant difference in survival between groups. The platform requires minimum 10 samples per group to calculate log-rank test statistics.

Mutation Frequency Analysis

Calculation of mutation prevalence across samples and genes. Includes identification of significantly mutated genes and hotspot mutations within protein domains.