Section: New Results

Sequence and structure annotation

Participants : Catherine Belleannée, François Coste, Jacques Nicolas.

Better scoring schemes for the recognition of functional proteins by protomata The machine learning algorithm included in Protomata-learner learns weighted automata representing both functional families from the sequences of amino acids, and the possible disjunctions between members. We investigated alternative sequence weighting strategies and null-models. We introduced a normalization of the score, and a method to assess the significance of scores, to simplify the prediction. Preliminary results show a good improvement of the prediction power of the computed models. [F. Coste] [36]

Detection of mutated primers and impact on targeted metagenomics results In targeted metagenomics, an initial task is the detection in each sequence of the primers used for amplifying the targeted region. The selected sequences are then trimmed and clustered in order to inventory the species present in the sample. Common pratices consist in retaining only the sequences with perfect primers (i.e. non-mutated by sequencing error). In the context of a study characterizing the biodiversity of tropical soils in unicellular eukaryotes, we have implemented the search for mutated primers, using the grammatical pattern matching tool Logol, and shown that retrieving sequences with mutated primers has a significant impact on targeted metagenomics results, as it makes possible to detect more species (7% additional OTUs in our study). [C. Belleannée] [34].

First landscape of binding to chromosomes for a domesticated mariner transposase in the human genome. In order to study the diversity of genomic targets of the SETMAR protein in two colorectal cell lines, a first task was to massively discover the Made1 80-bp transposon element in the human genome. For that, we used our Logol grammar-like approach to look for non perfect Made1 instances. In Logol, a pattern can be divided into several sub-patterns. The Made1 model took advantage of this feature to strengthen the most conserved regions. Cumulating this search with the Blast alignment search permitted to significantly increase the Made1 annotation in the human genome.[C. Belleannée] [33]