README.md

RNA Sequencing Project HS2022

This is the code used for the RNA sequencing project where the goal was to detect differentially expressed genes in breast cancer subtypes based on data of the paper "Transcriptomic landscape of breast cancers through mRNA sequencing" by Eswaran et al. (2012, https://doi.org/10.1038/srep00264).

The script for step 1-3 of the analysis can be found in the bash_scripts folder, the subsequent analysis with R in the rna_seq_project/rna_seq_analysis_R/ folder.

Steps of the analysis:

1. Quality Control

For the quality control, the rna_seq_project/bash_scripts/quality_checks.sh as well as the rna_seq_project/bash_scripts/trimmomatic.sh script was used. To improve the quality of individual samples, trimmomatic was used in paired end mode (PE option). This was only necessary for the second sample of the HER2 subgroup. The bases at the beginning and end of a read were cut at a quality threshold of 3 (LEADING:3 and TRAILING:3). The reads were scanned with a sliding window of 4 bases and cut when the average quality per base dropped below a threshold of 15 (SLIDINGWINDOW:4:15). The chosen minimum length was 40 bases (MINLEN:40).

2. Mapping reads to the reference genome

The reference genome used for this step ("Homo_sapiens.GRCh38.dna.primary_assembly.fa.gz") can be found at https://ftp.ensembl.org/pub/release-108/fasta/homo_sapiens/dna/, the annotation file "Homo_sapiens.GRCh38.108.gtf.gz" at https://ftp.ensembl.org/pub/release-108/gtf/homo_sapiens/.

2.1 Production of Index files using Hisat2

rna_seq_project/bash_scripts/produce_index_files_hisat2.sh

2.2 Mapping of reads to reference genome for each sample separately

rna_seq_project/bash_scripts/map_reads_to_genome_hisat2.sh (-1: Forward strand, R1, -2: reverse strand, R2)

2.3 Convert the resulting sam files to bam format using Samtools

rna_seq_project/bash_scripts/convert_sam_to_bam_samtools.sh (include header in output (-h), give output as BAM format (-b).

2.4 Sort the bam files by genomic coordinates using Samtools

rna_seq_project/bash_scripts/sort_bam_files_samtools.sh (-m <memory> -> 3500M)

2.5 Index the coordinate sorted bam files using Samtools

rna_seq_project/bash_scripts/index_samtools.sh

3. Counting the number of genes using FeatureCounts

rna_seq_project/bash_scripts/feature_counts.sh. Option '-p' is used for paired data. The -C option excludes read pairings that did not map correctly relative to one another. The strandedness can be defined using the -s option (0: unstranded). The programme expects name-sorted bam files as input, but if you submit coordinate-sorted bam files instead, featureCounts will resort them on the fly. The minimum filter threshold for mapping quality (-Q option) was set to 25. Make sure to provide several cores with the '-T' option if you want to use multiple cores. Since feature counts was run on each sample separately, the count matrix had to be created using the bioinfokit python module (rna_seq_project/bash_scripts/generate_count_matrix.py).

4. Exploratory data and differential expression analysis using DESeq2

rna_seq_project/rna_seq_analysis_R/diff_expression_analysis_final.R

5. Overrepresentation analysis using ClusterProfiler

rna_seq_project/rna_seq_analysis_R/overrepresentation_analysis_final.R