*2.2. Construction of cDNA Library and Gene Mapping to the Reference Genomes*

Four cDNA libraries were constructed using Illumina Hiseq 2000 platform (Illumina, San Diego, CA, USA) and 44,632,028, 45,110,536, 45,380,086, 44,434,116 high-quality reads were obtained, respectively (Table S1 in supplememntary materials). The sequencing raw data has been deposited into the Short Reads Archive (SRA) database under the accession number SRP166213. These clean reads were mapped to reference genome of *Phalaenopsis equestris* and the average gene mapping ratio of each sample was 55.47%. We considered that the low homology between *Phalaenopsis* 'Panda' and *Phalaenopsis equestris* result to the low mapping ratio, but it has no effect on RNAseq quantitative analysis owing to the high clean reads quantity and enough sequencing data (Table S2).

### *2.3. Functional Annotation and Classification*

To annotate the gene with putative functions, the assembled genes were searched against the public databases of NR. Among them, 17,871 genes were annotated to the NR database. To further illustrate the main biological functions of the transcripts, GO (15,588 genes) and KEGG pathway (19,864 genes) analyses were performed (Table S3).
