Whole Genome Assembly and Annotation for M. beccarii

Gene prediction was first performed by LoReAn (Cook et al. 2019), an automated annotation pipeline designed for eukaryotic genome annotation. Besides ab initio gene prediction, in this predicting pipeline, both long and short RNA-seq reads and protein sequences from three species, M. balbisiana, M. schizocarpa, M. acuminata, were included for RNA-seq and protein evidence-based gene prediction.

The LoReAn results were then input into funannotate pipeline v1.7 (https://github.com/nextgenusfs/funannotate) to obtain final integrated and consensus gene sets using the command of “funannotate train” and “funannotate predict”. During funannotate prediction, we applied the parameters of “-max_intronlen 100,000 -busco_db embryophyta -organism other”. After gene prediction, the command “funannotate annotate” was used for gene functional annotation. The annotation databases used included dbCAN v9.0 (Zhang et al. 2018), eggNOG v5.0.2 (Huerta-Cepas et al. 2017), GO (Gene Ontology) (Ashburner et al. 2000; The Gene Ontology Consortium 2019), KEGG (Kanehisa et al. 2016), InterPro v5.52-86 (Mitchell et al. 2019), MEROPS v12.2 (Rawlings et al. 2017), Pfam v34.0 (El-Gebali et al. 2019), and UniProt v2021_03 (The UniProt Consortium 2019).

For gene function comparison, the genes in E. glaucum, M. balbisiana, M. itinerans, M. schizocarpa and M. acuminate were also functional annotation using the command “funannotate annotate” as M. beccarii.