RNA-seq analysis

1. All analysis up to step 5 was performed in a Linux environment with custom bash scripts.
2. Adapter sequences were removed from raw sequencing reads using cutadapt (v1.13), with a command as in the example below:
   
   cutadapt -b universal_adapter –b index_adapter -o Trim_sample_R1.fastq.gz sample_R1.fastq.gz
   
   Illumina TrueSeq adapters were used. The universal adapter used was 'AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT'. The indexed adapter was specific based on each sample’s bar code.
    
3. Trimmed reads were then aligned to reference genome GRCm38 (GENCODE vM15 Primary assembly) using STAR aligner (Dobin et al., 2013) (version 2.5.3a). Index generation and alignment were both run using default parameters with commands as in the examples below:
   
   STAR --runThreadN 8 --runMode genomeGenerate -- genomeDir myGenomeDir --genomeFastaFiles myGenomeFile_fa --sjdbGTFfile myTranscriptsFile_gtf --sjdbOverhang 99
   
   STAR --runThreadN 10 --genomeDir myGenomeDir --outSAMtype BAM SortedByCoordinate --alignSJoverhangMin 8 --alignSJDBoverhangMin 1 --quantMode GeneCounts --outFileNamePrefix sample --readFilesCommand gunzip -c --readFilesIn Trim_sample_R1.fastq.gz Trim_sample_R2.fastq.gz
    
4. The number of reads aligned within the gene body (from TSS to TES, coordinates used from GENCODE vM15 transcripts annotation) of each gene was tabulated using FeatureCount (Liao et al., 2014) (v1.5.3) (without extension on both ends). FeatureCount was run with a command as in the example below:
   
   featureCounts -a myAnnotationFile_saf -F SAF -O -s 0 -p -B -C -T 5 -o sample_FCounts_NExt.txt sampleAligned.sortedByCoord.out.bam 
    
5. Differential gene expression (DEG) analyses on the read counts were performed using DESeq2 (Love et al., 2014) (v1.6.2) in R environment. Genes with total read counts less than 10 were filtered out from analysis. Design matrix was constructed with condition factor of three levels: WT, Dnmt3acKO, and Mecp2cKO. Results were generated from both “condition_Dnmt3a cKO_vs_WT” and “condition_Mecp2 cKO_vs_WT” contrasts.
6. For analysis of DEGs split according to gene length as measured from transcription start site to transcription end site (Coordinates used are from GENCODE vM15 transcripts annotation).  Genes were defined as long or short according to length standards defined in Gabel et al. (2015).

Category