• classification
• ferns
• identification
• LTR
• Polypodiopsida
• repetitive component
• retrotransposons

Genome divergence through gain and/or loss of the repetitive component plays a crucial role in the evolution of species. Nevertheless, knowledge of the variability associated with repeat proliferation in a wide range of plant species is still limited. Only partial data are available for species belonging to the class Polypodiopsida, commonly known as ferns. This thesis aims to fill these gaps by providing a comprehensive survey of long terminal repeat (LTR) retrotransposons in fern species. The focus is on the exhaustive identification and classification of these elements in six different species of Polypodiopsida (Ceratopteris richardii, Adiantum capillus-veneris, Alsophila spinulosa, Marsilea vestita, Salvinia cucullata, and Azolla filiculoides). For the identification and classification of fern LTR retrotransposons, a combination of structure-based and similarity-based strategies supported by an iterative approach was used. This method proved to be effective in the identification and annotation of repeated elements. The strategy was implemented using the REXdb database, a database of annotated LTR retrotransposon protein domains, and its associated tools DANTE and DANTE_LTR, widely recognized for their reliability and accuracy.
To achieve this, elements identified by their structural features but not classified were selected. These elements were classified with the aid of a phylogenetic tree constructed from protein sequences representing previously annotated elements from REXdb. The newly classified LTR retrotransposons were then added to REXdb as known elements, and the process was repeated. Adding multiple elements with the phylogenetic annotation significantly improved the detection and classification of additional LTR retrotransposons. This iterative approach resulted in a sevenfold increase in the number of identified instances compared to the first round of analysis, from 40197 elements in the first analysis to 279995 at the end of the recursive process. Furthermore, eight previously undetected lineages were identified, three of which are potentially new lineages.
A preliminary analysis of the genomic proportion of LTR retrotransposons revealed that in the genomes of heterosporous fern species, which are smaller, LTR retrotransposons account for only 10 to 40% of the genome. In contrast, in species with larger genomes, which are typical for homosporous ferns, LTR retrotransposons can constitute more than 50% of the genome. Finally, the evolutionary dynamics of fern LTR retrotransposons were investigated by comparing them with known repeat lineages present in other orders, such as flowering plants.