Supplementary Figure 3: Improved transcriptome analyses using the W22v2 genome.

a, RNA-seq data derived from endosperm tissue of W22 (SRA: SRR1986376) were aligned to both B73v4 and W22v2 using the same parameters. The percentage of mapped reads improves by mapping the data to W22. b, A set of 20,994 syntenic orthologous genes in B73 and W22 were identified and used for comparison of expression levels in alignments to B73 or W22. A comparison of expression level (reads per million, RPM) shows generally similar estimates in both genotypes with some genes that have differing expression estimates depending upon which genome was used for a reference. c–f, Differences in RPKM estimates can result from differences in annotation or mapping efficiency. c,d, Overlapping gene models in B73 result in all reads mapping to Zm00001d028756 (orange transcript) to be called ambiguous, while the corresponding gene in W22, Zm00004b001227 (orange transcript), has reads assigned. The adjacent gene, Zm00004b001228 in W22 and Zm00001d028757 in B73 (teal transcripts), has the same number of reads assigned to each reference, with a lower RPKM value reported in B73 owing to the longer gene model. e,f, The RPKM value for genes Zm00004b005439 in W22 and Zm00001d034453 in B73 is higher when mapping to W22 (e) than to B73 (f) owing to improved alignment to several regions of the gene (marked in yellow). Mapped reads are colored by strand: blue, forward; red, reverse.