Mitochondrial genome encoded gene expression values from RNA sequencing data from multiple tissue types

We reprocessed 11,371 RNA-Seq libraries across 36 different tissue/cell types to generate expression values for 15 genes encoded in the mitochondrial genome. RNA sequencing reads were trimmed for adaptor sequences, terminal bases with nucleotide quality below 20 and poly(A) tails >4bp in length. We then mapped data to a reference genome (1000G GRCh37 reference, which contains the mitochondrial rCRS NC_012920.1) with STAR 2.51a, using two-pass mapping, and then then filter out non-properly paired and non-uniquely mapped reads. Parameters and filtering steps were designed to improve alignment to the mitochondrial genome. We then calculated transcripts per million (TPM) per gene, before log10 normalizing expression values and removing outlier values. A description of the full pipeline, methods and data can be found in "Nuclear Genetic Regulation of the Human Mitochondrial Transcriptome" by Ali et al (2019) [PMID: 30775970].