Differential Transcript Profiling via RNA-seq

Sequencing data was demultiplexed to generate FASTQ files using Illumina bcl2fastq2 Conversion Software. FASTQ files were assessed with FastQC^{22 ,23} to verify that there was high sequence quality, expected sequence length, and no adapter contamination. Paired-end FASTQ files for each replicate were mapped to the Ensembl human reference transcriptome (GRCh38)^{24 (link)} using Kallisto^{25 (link),26} . Abundance data generated with Kallisto was read into R, annotated with Ensembl human gene annotation data (version 86)^{24 (link)}, and summarized as log2 counts per million (cpm) at the transcript level. Transcripts with greater than 0.2488206 cpm in at least 3 samples were retained after filtering. This threshold was chosen because it selected for transcripts with at least 10 counts in the smallest library sample. Samples were normalized with the trimmed mean of M values (TMM) method^{27 (link)}. The R package ‘limma’^{28 (link)} was used to identify differentially expressed transcripts by first applying precision weights to each transcript based on its mean-variance relationship using the VOOM function and then linear modeling and Bayesian statistics were employed to detect transcripts that were up- or down-regulated in each condition. Transcripts with an adjusted P-value < 0.05 and |log2 fold change| > 1 were considered significantly differentially expressed.