Long terminal repeat (LTR) retrotransposons are mobile elements that constitute the major fraction of most plant genomes. The identification and annotation of these elements via bioinformatics approaches represent a major challenge in the era of massive plant genome sequencing. In addition to their involvement in genome size variation, LTR retrotransposons are also associated with the function and structure of different chromosomal regions and can alter the function of coding regions, among others. Several sequence databases of plant LTR retrotransposons are available for public access, such as PGSB and RepetDB, or restricted access such as Repbase. Although these databases are useful to identify LTR-RTs in new genomes by similarity, the elements of these databases are not fully classified to the lineage (also called family) level. Here, we present InpactorDB, a semi-curated dataset composed of 130,439 elements from 195 plant genomes (belonging to 108 plant species) classified to the lineage level. This dataset has been used to train two deep neural networks (i.e., one fully connected and one convolutional) for the rapid classification of these elements. In lineage-level classification approaches, we obtain up to 98% performance, indicated by the F1-score, precision and recall scores.
Coffee leaf rust is the most damaging disease for coffee cultivation around the world. It is caused by a fungal pathogen, Hemileia vastatrix (Hva), belonging to the phylum Basidiomycota. Coffee leaf rust causes significant yield losses and increases costs related to its control, with evaluated losses of USD 1–2 billion annually. It attacks both the cultivated coffee species Coffea canephora (Robusta coffee) and Coffea arabica (Arabica coffee). New races, or pathotypes, are constantly emerging with increased virulence, suggesting a rapid evolution of the pathogen. Previous genetic and genomic studies have indicated a limited nucleotide diversity of Hva despite a high genetic diversity and large genome size estimated to be ~800 Mb, with a high content of repeated sequences (>74%). Despite several genomic resources and the release of a recent partial genome sequence, the diversity of these repeated sequences and how they may impact the evolution of the leaf rust genome have not been investigated in detail. In an attempt to characterize the transposable elements within the Hva genomes, we report here new lineages of long terminal repeat (LTR) retrotransposons, called CO-HUI, Soroa, and Baco, which are classified into Gypsy, and and Labe and Mapi, which are classified as Copia. The CO-HUI and Soroa elements represent the main part of all Hva transposable elements, as well as approximately 37% of the available genome assemblies. Mapi and CO-HUI are the main expressed families in RNA-seq data. Although Soroa is the lineage showing more insertions into exons and genes, Mapi seems to be more frequently involved in co-expression with genes. All these new families are also present in the Pucciniales, suggesting that they dynamically participate in their genome evolution.