We quantified pluripotency (or lack thereof, indicating successful differentiation) of iPSCs using differentiation marker expression levels as previously described (Carcamo-Orive et al., 2016). Of the total 24 differentiation markers, we used 16 markers with minimum estimated three log2 fold change (FC) across samples. Previously published human iPSC and ESC gene expression quantifications (Choi et al., 2013) were included as controls. We downloaded relevant TPM quantification matrices from the Gene Expression Omnibus (GEO) database (GEO Accession: GSE 73211). TPM quantifications from our cohorts and those of Choi et al. were quantile normalized per tissue, log2 transformed to estimate FC, and then normalized with respect to each other as a merged dataset. Eigenvalues were estimated using singular value decomposition (SVD) as implemented in the base R function svd() for the centered dataset. Variance contributed by each component was estimated as D2/ΣD2.
