Translation arrest directed by nascent peptides and small cofactors controls expression of important bacterial and eukaryotic genes, including antibiotic resistance genes, activated by binding of macrolide drugs to the ribosome. Previous studies suggested that specific interactions between the nascent peptide and the antibiotic in the ribosomal exit tunnel play a central role in triggering ribosome stalling. However, here we show that macrolides arrest translation of the truncated ErmDL regulatory peptide when the nascent chain is only three amino acids and therefore is too short to be juxtaposed with the antibiotic. Biochemical probing and molecular dynamics simulations of erythromycin-bound ribosomes showed that the antibiotic in the tunnel allosterically alters the properties of the catalytic center, thereby predisposing the ribosome for halting translation of specific sequences. Our findings offer a new view on the role of small cofactors in the mechanism of translation arrest and reveal an allosteric link between the tunnel and the catalytic center of the ribosome.

We have developed an easy-to-use and memory-efficient method called PhenotypeSeeker that (a) identifies phenotype-specific k-mers, (b) generates a k-mer-based statistical model for predicting a given phenotype and (c) predicts the phenotype from the sequencing data of a given bacterial isolate. The method was validated on 167 Klebsiella pneumoniae isolates (virulence), 200 Pseudomonas aeruginosa isolates (ciprofloxacin resistance) and 459 Clostridium difficile isolates (azithromycin resistance). The phenotype prediction models trained from these datasets obtained the F1-measure of 0.88 on the K. pneumoniae test set, 0.88 on the P. aeruginosa test set and 0.97 on the C. difficile test set. The F1-measures were the same for assembled sequences and raw sequencing data; however, building the model from assembled genomes is significantly faster. On these datasets, the model building on a mid-range Linux server takes approximately 3 to 5 hours per phenotype if assembled genomes are used and 10 hours per phenotype if raw sequencing data are used. The phenotype prediction from assembled genomes takes less than one second per isolate. Thus, PhenotypeSeeker should be well-suited for predicting phenotypes from large sequencing datasets. PhenotypeSeeker is implemented in Python programming language, is open-source software and is available at GitHub (https://github.com/bioinfo-ut/PhenotypeSeeker/).

ConoDictor is a tool that enables fast and accurate classification of conopeptides into superfamilies based on their amino acid sequence. ConoDictor combines predictions from two complementary approaches—profile hidden Markov models and generalized profiles. Results appear in a browser as tables that can be downloaded in various formats. This application is particularly valuable in view of the exponentially increasing number of conopeptides that are being identified. ConoDictor was written in Perl using the common gateway interface module with a php submission page. Sequence matching is performed with hmmsearch from HMMER 3 and ps_scan.pl from the pftools 2.3 package. ConoDictor is freely accessible at http://conco.ebc.ee.