Clinical Genomics Project: Whole Exome Sequence Analysis of Ewing Sarcoma Sample

Overview

This project focuses on the analysis of Whole Exome Sequencing (WES) data from a tumor tissue sample of a 20-year-old male patient diagnosed with Ewing Sarcoma. The goal is to identify and annotate genetic variants within the protein-coding regions (exome) of the human genome to understand mutational impacts and potential clinical relevance.

Whole Exome Sequencing targets the exonic regions which harbor approximately 85% of disease-associated variants, making it an essential tool in cancer genomics and precision medicine.

---

Project Details

Experimental Design

• The project analyzes WES data retrieved from the Sequence Read Archive (SRA).
• Sample: BioSample ID SAMN35991112, SRA ID SRS18070572
• BioProject: PRJNA987736 titled “Immunosurveillance shapes the emergence of neo-epitope landscapes of sarcomas, revealing prime targets for immunotherapy”.
• Raw reads were downloaded from SRA, quality assessed, trimmed, aligned to human reference genome (hg38), and variants were called, filtered, and annotated.

Sample Information

Field	Value
Organism	Homo sapiens
Sample Source	Tumor tissue of 20-year-old male patient
Collection Date	December 18, 2019
Location	Pittsburgh, USA
Sequencing Platform	Illumina NovaSeq 6000
Read Count	22,414,250 paired-end reads
Base Count	6,724,275,000 bases
Library Strategy	Whole Exome Sequencing (WXS)
Library Preparation	DNA fragmentation by sonication, PCR enrichment
Data Package Version	Human data package v1.0

---

Methodology

1. Data Collection

• Raw paired-end reads were downloaded from SRA using an automated shell script.
• Quality criteria for dataset selection: minimum file size ≥ 2GB and Phred quality score ≥ 30.

2. Quality Control (QC)

• FastQC was used to assess raw data quality, analyzing metrics such as:
  • Per base sequence quality
  • GC content
  • Adapter contamination
  • Sequence duplication levels
• QC results guide the necessity for further preprocessing.

3. Pre-Processing

• Fastp trimmed adapter sequences, removed low-quality and redundant reads.
• Post-trimming quality was verified again with FastQC.

4. Alignment

• Cleaned reads aligned to the human reference genome (hg38) using BWA.
• Reference genome was indexed with BWA prior to alignment.
• SAM files generated and converted to BAM format using samtools.

5. Post-Alignment Processing

• BAM files sorted (GATK SortSam) and duplicates marked/removed (GATK MarkDuplicates).
• Alignment metrics collected using GATK CollectAlignmentSummaryMetrics.
• BAM files indexed with samtools index.
• Base Quality Score Recalibration (BQSR) performed using GATK, leveraging known variant datasets to improve base call accuracy.

6. Variant Calling

• Variants called using GATK HaplotypeCaller generating VCF files.

7. Variant Filtering

• Hard filtering applied with GATK VariantFiltration using quality thresholds:

Filter	Condition	Purpose
QD2	QD < 2.0	Low quality relative to depth
QUAL30	QUAL < 30.0	Low variant confidence
SOR3	SOR > 3.0	Strand bias
FS60	FS > 60.0	Fisher Strand bias
MQ40	MQ < 40.0	Low mapping quality

• Variants passing all filters (“PASS”) were retained using bcftools.

8. Variant Quality Score Recalibration (VQSR)

• Applied machine learning based recalibration separately for INDELs and SNPs.
• Training datasets included Mills, Axiom Exome Plus, HapMap, Omni, and 1000 Genomes.
• Final high-confidence variants extracted and separated into SNPs and INDELs using GATK SelectVariants.

9. Variant Annotation

• SnpEff annotated filtered variants against the hg38 reference, predicting functional effects on genes.
• Further detailed biological context was obtained using Variant Effect Predictor (VEP) from Ensembl, providing consequence types, gene names, dbSNP IDs, and clinical significance.
• VEP results exported to Excel for convenient analysis.

---

Tools and Software

Step	Tool/Software
Quality Control	FastQC
Preprocessing	Fastp
Alignment	BWA, samtools
BAM Processing	GATK (SortSam, MarkDuplicates, CollectAlignmentSummaryMetrics, BaseRecalibrator)
Variant Calling	GATK HaplotypeCaller
Variant Filtering	GATK VariantFiltration, bcftools
Variant Recalibration	GATK VariantRecalibrator, ApplyVQSR
Annotation	SnpEff, Variant Effect Predictor (VEP)

---

Data Structure

/data/
├── raw/ # Raw sequencing FASTQ files
├── reference/ # Reference genome and annotation files (hg38)
/results/
├── qc_reports/ # FastQC reports
├── trimmed_reads/ # Cleaned FASTQ files
├── alignments/ # BAM and related files
├── variants/ # VCF files (raw, filtered, recalibrated, annotated)
└── annotation/ # SnpEff and VEP output files

---

#References

BioProject PRJNA987736: https://www.ncbi.nlm.nih.gov/bioproject/PRJNA987736

SRA Database: https://www.ncbi.nlm.nih.gov/sra

GATK Best Practices: https://gatk.broadinstitute.org/

SnpEff: http://snpeff.sourceforge.net/

Variant Effect Predictor (VEP): https://www.ensembl.org/info/docs/tools/vep/index.html