*4.6. Expression Analysis of TPSs in Rosaceae*

For the three representative plants (*F. vesca*, *M. domestica*, *P. persica*), RNA-seq data of different tissues and development stages (ripe fruit, immature fruit, leaf) was retrieved from the fruitENCODE project [37]; each tissue consisted of two replicates. The fruitENCODE project aims to generate a comprehensive annotation of functional elements in seven climacteric fruit species (apple, banana, melon, papaya, peach, pear, and tomato) with sequenced reference genomes. The data were deposited in the SRA database of NCBI with accession number PRJNA381300. Transcriptome analysis was implemented by the protocol in a previous study [38]. The clean reads of RNA-seq data from each sample were mapped against the genome reference with HISAT2 [39]; each SAM file was converted into a BAM file, and sorted. Duplicates were removed with SAMtools [40,41]. Further transcript assembly and quantification of the read alignments were performed using Stringtie [42]. Gene expression levels were measured by FPKM (fragments per kilobase of transcript per million mapped reads) and normalized with the row-scale method. Heatmaps with all samples were plotted using the "HeatMap" function in TBtools.
