Skip to content

Introduction

Introduction

This bioinformatics pipeline is containerized and platform-independent, built using a Nextflow DSL2 workflow, and deployed on both the NIH Biowulf HPC cluster and the AWS cloud.

This Exome–RNA-seq workflow accepts a samplesheet and FASTQ files as input, then performs a comprehensive set of analyses, including:

  • Extensive quality control
  • Mutation calling
  • Tumor mutational burden assessment
  • Mutational signature analysis
  • HLA typing
  • Copy-number variation detection
  • T-cell infiltration prediction
  • Neoantigen prediction
  • Gene expression profiling
  • Fusion detection
  • Variant calling and annotation

This end-to-end suite supports deep genomic characterization for both research and institutional datasets at NIH.

Results can be explored via the ClinOmics Data Portal, which provides a secure, interactive, user friendly web interface for data visualization and interpretation. It is well suited for both individual patient level and cohort wide cancer genomics analysis.

Who This Is For

This workflow is intended for researchers and bioinformaticians working with exome and RNA-seq datasets who need a reproducible, scalable, and validated solution for comprehensive genomic analysis. It is designed for use in both HPC and cloud computing environments.

Here is a Snapshot of our RNAseq and Exome workflows.

RNAseq Workflow DAG

RNAseq Workflow DAG

Exome Workflow DAG

Exome_workflow

Workflow Tools

Stage Tool(s) Purpose
Alignment STAR Align RNA-seq reads to the reference genome with splice-aware mapping
Strandedness Assessment ngsderive Determine library strandedness from aligned BAM files
Expression Quantification RSEM Quantify gene and transcript abundance
Post-mapping QC Picard CollectRNASeqMetrics, RNA-SeQC Evaluate mapping quality, insert size, and gene body coverage
BAM Processing GATK SplitNCigarReads, RealignerTargetCreator, IndelRealigner, BaseRecalibrator, PrintReads Prepare RNA-seq BAMs for variant calling by handling spliced reads, realigning indels, and recalibrating base quality scores
Variant Calling GATK HaplotypeCaller Identify germline SNPs and indels from RNA-seq data
Fusion Detection STAR-Fusion, Arriba, FusionCatcher Detect gene fusions; integrate consensus calls across multiple tools
HLA Calling HLA-HD, OptiType Predict HLA types from RNA-seq data
Variant Annotation ANNOVAR Annotate variants with functional consequences and external database references
Stage Tool(s) Purpose
Alignment BWA-MEM Align exome reads to the reference genome
BAM QC Picard CollectAlignmentSummaryMetrics, Picard CollectInsertSizeMetrics Evaluate mapping quality, duplication rates, and insert size
BAM Processing GATK MarkDuplicates, BaseRecalibrator, ApplyBQSR Mark duplicates, recalibrate base quality scores, and prepare BAMs for variant calling
Germline Variant Calling GATK HaplotypeCaller Identify germline SNVs and indels
Somatic Variant Calling Manta, Strelka2, GATK Mutect2 Identify somatic SNVs and indels
Copy Number Calling CNVkit Detect copy number alterations
Variant Annotation ANNOVAR, SnpEff, VEP Annotate variants with functional consequences and external database references
T-cell Infiltration Estimation TcellExTRECT Estimate tumor-infiltrating T-cell abundance from sequencing data
Tumor Ploidy & Cellularity Sequenza Infer tumor purity and ploidy
Microsatellite Instability MANTIS Detect microsatellite instability from tumor/normal pairs
Mutational Signatures SigProfilerExtractor, COSMIC Signature Database Identify mutational signatures present in the tumor genome
HLA Typing HLA-HD, OptiType Predict HLA types from sequencing data
Neoantigen Prediction pVACseq Predict candidate neoantigens from tumor-specific variants
Concordance Analysis Conpair Verify tumor/normal sample pairing via genotype concordance
Stage Tool(s) Purpose
Quality Control FastQC, MultiQC Assess raw read quality and generate aggregated QC reports
Contamination Screening Kraken2, Krona, FastQ Screen Identify potential contamination from non-target species
Read Trimming Cutadapt Remove adapters, low-quality bases, and trim reads for improved alignment
BAM QC & Coverage Analysis samtools, bamUtils, mpileup, bedtools Generate hotspot coverage plots, transcript coverage plots, circos plots, and perform BAM compression/clean-up
QC Aggregation MultiQC Aggregate QC metrics across all steps into a single interactive report

Before You Begin

If you are setting up this workflow for the first time, ensure your environment meets all prerequisites and dependencies.
See the Prerequisites section below and If you are new to Nextflow, please refer to this page for quick introduction.

Prerequisites

To run this workflow, you will need the following software:

    Nextflow >= 21.04.3
    Singularity 3.10.5
    Graphviz 2.40

Architecture

  1. The workflow is located in the Khanlab space: /data/khanlab/projects/Nextflow_dev/dev/AWS_POC_MVP_NF.

  2. The biowulf_nextflow.config file in this directory contains the master configuration. A config/ subfolder contains additional configuration files:

    config/
├── aws_cluster_test.config
├── aws_ec2_test.config
├── aws_params.config
├── biowulf_cluster_test.config
├── biowulf_s3_test_params.config
├── biowulf_singularity.config
├── biowulf_small_test_params.config
├── biowulf_small_test_params_hg38.config
├── docker.config
├── GRCm39.config
└── omics.config

  3. Pipeline can be launched using the script launch.sh. It takes samplesheet.csv as a mandatory input.

  4. Guidelines to create an input samplesheet can be found here.
  5. All the references, annotation and bed files are currently located under /data/khanlab. We currently support data processing for these capture kits.

Sequencing capture kits

RNAseq Exome
Access clin_ex_v1
PolyA seqcapez.hu.ex.v3
PolyA_stranded Agilent_v7
Ribozero idt_v2_plus
smartrna xgen_hyb_panelv2
seqcapez_hu_ex_v3

If you want to process the data sequenced by other kits, please reach out to Vineela Gangalapudi.