This presentation introduces the multidisciplinary field of bioinformatics, detailing its integration of biology, computer science, statistics, mathematics, biochemistry, and chemistry to analyze biological data and understand complex systems. It then focuses on a targeted enrichment analysis workflow for germline variants using whole-exome sequencing data. The speaker outlines the comprehensive process from DNA extraction and automated library preparation (using Kapa HyperExome V2 and Avenio Millis system) through to sequencing on Illumina NovaSeq. The core of the presentation dives deep into the bioinformatics pipeline, covering primary analysis (BCL to FASTQ conversion, QC, trimming, optional downsampling), secondary analysis (read mapping, duplicate removal, coverage statistics, variant calling with DeepVariant), and tertiary analysis (variant annotation with SnpEff and filtering with SnpSift). The presentation concludes with a summary of results, emphasizing the pipeline's ability to detect germline SNVs and small InDels, and notes that the workflow has been wrapped in a containerized solution for streamlined analysis.
Summarizing the Bioinformatics Targeted Enrichment Workflow
0:05 Introduction to Bioinformatics: Bioinformatics is a complex and exciting field that integrates biology, computer science, statistics, mathematics, biochemistry, and chemistry to analyze DNA, RNA, proteins, and understand biological systems.
0:40 Diverse Disciplines: Within biology, bioinformatics encompasses genomics (DNA structure/function), epigenomics (gene expression), transcriptomics (RNA transcripts), proteomics (protein structure), and metabolomics (small molecules), often utilizing Next-Generation Sequencing (NGS) data.
2:40 Targeted Enrichment Focus: The presentation focuses on targeted enrichment studies, such as whole-exome sequencing, as a cost-effective and precise alternative to whole genome sequencing for deciphering individual gene roles or investigating rare genetic variants.
3:14 Laboratory Workflow: The process begins with DNA extraction, followed by library preparation (fragmentation, end repair, A-tailing, adapter ligation, amplification), target enrichment, and sequencing using an Illumina NovaSeq instrument.
3:50 Automated Library Prep: The Kapa HyperExome V2 kit is used for genomic DNA library preparation, and the Avenio Millis system provides fully automated, end-to-end library preparation, target enrichment, quantification, normalization, and pooling.
4:40 Bioinformatics Pipeline Overview: Post-sequencing data undergoes a multi-stage bioinformatics analysis, including primary (data conversion, QC), secondary (alignment, variant calling), and tertiary (variant annotation and filtering) analyses.
4:49 Primary Analysis (BCL to FASTQ): Illumina's binary base call (BCL) files are demultiplexed and converted into text-based FASTQ files (containing raw sequence data and quality scores) using Illumina's BCL Convert tool.
5:33 Data Preprocessing: FASTQ files undergo quality control (QC), adapter trimming, and quality filtering using tools like Fastp. Optional downsampling (e.g., to 60 million reads using GATK) is performed for sample comparability and coverage normalization.
6:04 Secondary Analysis (Alignment & Duplicates): Reads are mapped to a reference genome (GRCh38/hg38) using BWA-MEM, generating BAM files. GATK tools like FixMates clean paired-read information, and duplicate reads (arising from PCR or over-sequencing) are marked and/or removed to prevent bias in variant calling.
13:39 Coverage & Capture Statistics: Key metrics calculated include a low duplication rate, 100% mapping rate, low alignment error rate, ~90% of targets with at least 30X coverage depth, and a low fold-80 base penalty indicating uniform coverage. The on-target rate is around 80%.
15:20 Variant Calling & Annotation: DeepVariant, a deep learning-based variant caller, identifies single nucleotide variants (SNVs) and insertions/deletions (InDels) and reports them in VCF format. SnpEff annotates these variants for predicted impact (high, medium, low), and SnpSift filters variants based on quality scores and impact.
18:01 Quality Control of Variants: The transition-to-transversion (Ts/Tv) ratio is monitored (expected ~2.8 for whole exome). A low ratio can indicate artifactual variants or false positives.
18:41 Key Takeaways: Bioinformatics is highly interdisciplinary, and numerous open-source tools facilitate pipeline construction. Targeted enrichment workflows offer sensitive and precise detection of germline SNVs and small InDels. The workflow described has been containerized for streamlined, repeatable analysis of Kappa TE datasets.