A self-attention-based deep learning model for identifying key genes in insect pupal metamorphosis

The protein sequences of the 293 insect species in the GNCM_293 insect.faa.zip file are the longest transcripts for each protein retained in this study.The protein sequences of 293 insects were obtained from National Center for Biotechnology Information ( https://www.ncbi.nlm.nih.gov/ ) , with protein-coding genes as the raw data samples (S1 Table). These insects were classified into two categories according to their metamorphic type characteristics (Hemimetabolous and Holometabolous): 36 hemimetabolous insects in 8 orders and 257 holometabolous insects in 6 orders, to study the functions of key genes for insect pupal metamorphic.These protein sequences were used to construct the GNCM and analyze key genes for insect metamorphosis.