2.5. TransDecoder Analysis

TransDecoder analysis (default parameters) was carried out according to the method described in [28]. Firstly, a minimum length of opening reading frame (ORF) (default 100 aa) was identified in the transcripts, to further maximize the sensitivity of the functional ORF. The ORF was compared with the UniProt and PFAM databases to identify the common protein domain. Finally, the final prediction was scored based on the alignment in the two databases. Based on the TransDecoder analysis, the transcripts were classified into four categories: complete, 5prime_partial, 3prime_partial, and internal. Complete means the sequence contains the complete opening reading frame (ORF), 5prime_partial means the 5′ end of the sequence was missing, 3prime_partial means the 3′ end was missing, and internal means both 5′ and 3′ parts were missing. For genes which were predicted to have multiple ORFs, the one with the highest score was reserved. For genes which were predicted as complete, 5prime_partial, or 3prime_partial, if the score was less than −20, they were merged with the internal category and renamed ‘internal and others’.