Whole Genome Assembly and Annotation of M. troglodytarum

The repeats and protein-coding genes in the genome were annotated by MAKER2 (v3.01.02). The MAKER2 pipeline was processed twice to obtain high-quality gene annotations. The RNA-seq reads of the leaves, stems, roots and fruits were imported to Trinity to generate genome-guided and de novo assemblies, with the default parameters used.

Then, the assemblies were imported into the PASA pipeline (v2.3.3) to construct a comprehensive transcript database. The comprehensive transcripts were used to train the predictors embedded in the MAKER2 pipeline, including AUGUSTUS (v3.3.1), GENEMARK (v3.5.2) and SNAP (version 2006-07-28).

After filtering out proteins produced by MAKER2 that had low AED values, the ab initio predictors AUGUSTUS, GENEMARK and SNAP were trained again.

Then, utilizing high-confidence and high-sensitivity transcripts produced by StringTie (v2.1.1) as inputs, the MAKER2 pipeline was run again. Using eggNOG-mapper (v2) and its default database, genes were assigned functional annotations.