RNA-Seq Analysis Pipeline

This pipeline uses an annotated genome to identify differential expressed genes/transcripts. 15 hour minimum ($1470 internal, $1860 external) per project.

1. Quality Assessment

Quality of data assessed by FastQC; results of quality assessment will be evaluated prior to downstream analysis.

Deliverables:
- reports generated by FastQC
Tools used:
- FastQC: (Andrews 2010) used to generate quality summaries of data:
  - Per base sequence quality report: useful for deciding if trimming necessary.
  - Sequence duplication levels: evaluation of library complexity. Higher levels of sequence duplication may be expected for high coverage RNAseq data.
  - Overrepresented sequences: evaluation of adapter contamination.

2. Fastq Preprocessing

Quality assessment used to decide if any preprocessing of the raw data is required and if so, preprocessing is performed.

Deliverables:
- Trimmed/filtered fastq files.
Tools Used:
- Fastx-toolkit: Used to preprocess fastq files.
  - Fastq quality trimmer: Trimming reads based on quality.
  - Fastq quality filter: Filtering reads based on quality.
- Cutadapt: Used to remove adaptor from reads.

3. Mapping

Mapping to transcriptome reference performed using Kallisto pseudomapper or mapping to genome reference performed using HISAT2.

Deliverables:
- Mapping results, as bam files (when mapped using HISAT2) and mapping statistics.
Tools Used:
- Kallisto: (Bray 2016) pseudoaligner and RNA-Seq quantification tool
- HISAT2: (Kim 2015) aligner used to generate read alignments in a splice-aware manner and identify novel junctions.
- Samtools: (Li 2009) used to generate mapping statistics.

4. Gene/Transcript Counting

Counting the number of reads mapping to annotated intervals to obtain abundance of genes/transcripts.

Deliverables:
- Raw gene/transcript counts
- Variance stabilized gene/transcript counts
Tools Used:
- Kallisto: (Bray 2016) pseudoaligner and RNA-Seq quantification tool
- HTSeq-count: (Anders 2014) used to count reads overlapping gene intervals.

5. DEG Identification

Normalization and statistical testing to identify differentially expressed genes.

Deliverables:
- DEG Summary and master file containing fold changes and p values for every gene.
Tools Used:
- DESeq2: (Love 2014) used to perform normalization and test for differential expression using the negative binomial distribution.

5. Visualizations

Standard visualizations of the RNA-Seq data using in-house R Scripts.

Deliverables:

Sample dendogram
Sample-Sample correlation plot
Pair plot: Matrix of scatter plots showing relationship of every sample metadata variable to every other variable.
Expression heatmap with clustering of samples
Volcano plot : Scatter plot of fold-change versus significance
Box plots of top 10 upregulated and top 10 downregulated genes.
PCA plot: Orthogonal transformation of the data to look at underlying structure of data.