A machine learning based framework to identify and classify long terminal repeat retrotransposons.

Transposable elements (TEs) are repetitive nucleotide sequences that make up a large portion of eukaryotic genomes. They can move and duplicate within a genome, increasing genome size and contributing to genetic diversity within and across species. Accurate identification and classification of TEs present in a genome is an important step towards understanding their effects on genes and their role in genome evolution. We introduce TE-Learner, a framework based on machine learning that automatically identifies TEs in a given genome and assigns a classification to them. We present an implementation of our framework towards LTR retrotransposons, a particular type of TEs characterized by having long terminal repeats (LTRs) at their boundaries. We evaluate the predictive performance of our framework on the well-annotated genomes of Drosophila melanogaster and Arabidopsis thaliana and we compare our results for three LTR retrotransposon superfamilies with the results of three widely used methods for TE identification or classification: RepeatMasker, Censor and LtrDigest. In contrast to these methods, TE-Learner is the first to incorporate machine learning techniques, outperforming these methods in terms of predictive performance, while able to learn models and make predictions efficiently. Moreover, we show that our method was able to identify TEs that none of the above method could find, and we investigated TE-Learner's predictions which did not correspond to an official annotation. It turns out that many of these predictions are in fact strongly homologous to a known TE.

On the Search for Retrotransposons: Alternative Protocols to Obtain Sequences to Learn Profile Hidden Markov Models.

Profile hidden Markov models (pHMMs) have been used to search for transposable elements (TEs) in genomes. For the learning of pHMMs aimed to search for TEs of the retrotransposon class, the conventional protocol is to use the whole internal nucleotide portions of these elements as representative sequences. To further explore the potential of pHMMs in such a search, we propose five alternative ways to obtain the sets of representative sequences of TEs other than the conventional protocol. In this study, we are interested in Bel-PAO, Copia, Gypsy, and DIRS superfamilies from the retrotransposon class. We compared the pHMMs of all six protocols. The test results show that, for each TE superfamily, the pHMMs of at least two of the proposed protocols performed better than the conventional one and that the number of correct predictions provided by the latter can be improved by considering together the results of one or more of the alternative protocols.

Learning HMMs for nucleotide sequences from amino acid alignments.

Profile hidden Markov models (profile HMMs) are known to efficiently predict whether an amino acid (AA) sequence belongs to a specific protein family. Profile HMMs can also be used to search for protein domains in genome sequences. In this case, HMMs are typically learned from AA sequences and then used to search on the six-frame translation of nucleotide (NT) sequences. However, this approach demands additional processing of the original data and search results. Here, we propose an alternative and more direct method which converts an AA alignment into an NT one, after which an NT-based HMM is trained to be applied directly on a genome.

Comparative analysis of differentially expressed sequence tags of sweet orange and mandarin infected with Xylella fastidiosa

The Citrus ESTs Sequencing Project (CitEST) conducted at Centro APTA Citros Sylvio Moreira/IAC has identified and catalogued ESTs representing a set of citrus genes expressed under relevant stress responses, including diseases such as citrus variegated chlorosis (CVC), caused by Xylella fastidiosa. All sweet orange (Citrus sinensis L. Osb.) varieties are susceptible to X. fastidiosa. On the other hand, mandarins (C. reticulata Blanco) are considered tolerant or resistant to the disease, although the bacterium can be sporadically detected within the trees, but no disease symptoms or economic losses are observed. To study their genetic responses to the presence of X. fastidiosa, we have compared EST libraries of leaf tissue of sweet orange Pêra IAC (highly susceptible cultivar to X. fastidiosa) and mandarin ‘Ponkan’ (tolerant) artificially infected with the bacterium. Using an in silico differential display, 172 genes were found to be significantly differentially expressed in such conditions. Sweet orange presented an increase in expression of photosynthesis related genes that could reveal a strategy to counterbalance a possible lower photosynthetic activity resulting from early effects of the bacterial colonization in affected plants. On the other hand, mandarin showed an active multi-component defense response against the bacterium similar to the non-host resistance pattern.

Identification and frequency of transposable elements in Eucalyptus

Transposable elements (TE) are major components of eukaryotic genomes and involved in cell regulation and organism evolution. We have analyzed 123,889 expressed sequence tags of the Eucalyptus Genome Project database and found 124 sequences representing 76 TE in 9 groups, of which copia, MuDR and FAR1 groups were the most abundant. The low amount of sequences of TE may reflect the high efficiency of repression of these elements, a process that is called TE silencing. Frequency of groups of TE in Eucalyptus libraries which were prepared with different tissues or physiologic conditions from seedlings or adult plants indicated that developing plants experience the expression of a much wider spectrum of TE groups than that seen in adult plants. These are preliminary results that identify the most relevant TE groups involved with Eucalyptus development, which is important for industrial wood production