Abstract:Genome-scale prediction of FOM（Fusarium oxysporum f.sp.melonis） secretome was firstly performed by combining different bioinformatic softwares,including SignalP,TMHMM,WOLFPRT,TargetP,big-PI Fungal Predictor,and SecretomeP in this study.Of the 26 811 FOM protein sequences,1 145 sequences (4.3%) were predicted to be classically secreted proteins and 8 471 sequences (31.6%) might be secreted in a non-classical manner.The 1 145 secreted proteins were characterized with 100 to 600 amino acids,and the length of the signal peptides was concentrated between 17 to 22 amino acids.Of the 1 145 classically secreted proteins,a total of 605 proteins were functionally annotated to be involved in carbohydrate metabolic process,transport process,hydrogen peroxide catabolic process and biosynthetic process.Among the 1 145 classically secreted proteins,277 proteins (24.2%) were predicted to be CAZymes.