Bulk RNA-seq and data analysis

Sequencing libraries from purified RNA were prepared with the QuantSeq 3’ mRNA-Seq Library Prep Kit FWD (Lexogen GmbH, Vienna, Austria). Samples were indexed to allow for multiplexing. Library quality and size range was assessed using 2100 Bioanalyzer with the DNA 1000 kit (both Agilent Technologies). The libraries were diluted to a final concentration of 2 nM and subsequently sequenced on an Illumina HiSeq4000 platform. Single-end reads of 50 bp length were produced with a minimum of 2M reads per sample. Quality control of raw reads was performed with FastQC v0.11.7 (37). Adapters were filtered with ea-utils fastq-mcf v1.05 (38). Using HiSAT2 (39), split-aware alignment was accomplished against the human reference genome hg19. Reads mapping to multiple loci in the reference genome were discarded. The resulting BAM files were handled with Samtools v1.5 (40). The reads per gene were quantified with HT-seq Count v2.7.14 (41). Count-based differential expression (DE) analysis was done with the R-based Bioconductor package DESeq2 version 1.34.0 (42). Reported p-values were adjusted for multiple testing with the Benjamini-Hochberg procedure (43), which controls the false discovery rate (FDR). Principal component analysis was used to inspect sample- and group-specific variation with the R package DESeq2 (42) using the top 500 most variable genes across all samples. Surrogate variable analysis with the Bioconductor package sva version 3.42.0 (44) was used to determine the age groups to perform relevant age adjustment in DE analysis. Raw sequencing data is available at the European Nucleotide Archive, accession no PRJEB50778.