The machine learning tool, psytrans was used to separate symbiont and host transcripts using the initial assembly produced by transfuse.
In order to learn the difference between host and symbiont psytrans requires reference databases. We use predicted transcripts from the Acropora millepora genome as a host reference and EST sequences from Breviolum minutum as a symbiont reference.
The output file species1_good.PdPf_transfuse95_cons.fa from psytrans represents the final reference transcriptome used in this study.
BUSCO results for this are;
C:97.0%[S:33.3%,D:63.7%],F:2.3%,M:0.7%,n:303
294 Complete BUSCOs (C) 101 Complete and single-copy BUSCOs (S) 193 Complete and duplicated BUSCOs (D) 7 Fragmented BUSCOs (F) 2 Missing BUSCOs (M) 303 Total BUSCO groups searched