2.3. Genome Component Prediction and Gene Annotation

Putative open reading frames (ORF) were predicted using the GeneMarkS v.4.17 program [23] (http://topaz.gatech.edu/GeneMark/). Transfer RNA (tRNA) genes were predicted with tRNAscan-SE [24]. Ribosome RNA (rRNA) genes were analyzed using rRNAmmer [25]. Small RNAs (sRNA) were predicted by BLAST against the Rfam database [26,27], and confirmed using the cmsearch program (version 1.1rc4) with default parameters. Gene annotations were determined with the BLASTP program (E-value < 1 ×10⁻⁵, identity ≥ 40%, coverage ≥ 40%) and six databases including GO (gene ontology) [28], KEGG (Kyoto encyclopedia of genes and genomes) [29,30], COG (clusters of orthologous groups) [31], NR (non-redundant protein databases) [32], TCDB (transporter classification database) [33], and Swiss-Prot [34]. Genome overview was created by Circos to show the annotation information [35].

The secretory proteins were predicted with SignalP (version 4.1, http://www.cbs.dtu.dk/services/SignalP-4.1/) [36] and TMHMM (Version 2.0c, http://www.mybiosoftware.com/tmhmm-2-0c-prediction-transmembrane-helices-proteins.html). Since X. albilineans is a bacterial pathogen, pathogenicity and drug resistance data were also retrieved from the pathogen–host interactions database (PHI-base) [37], the virulence factors of pathogenic bacteria database (VFDB) [38], and the antibiotic resistance genes database (ARDB) [39]. Carbohydrate-active enzymes were predicted using the carbohydrate-active enzymes database [40]. CRISPRs were identified using CRISPRdigger (version 1.0, https://github.com/greyspring/CRISPRdigger) [41].