Menú Accesibilidad

Logo of UPM

internacional

Palabra o palabras clave para realizar la búsqueda
Ingeniamos el futuros Acceder al menú accesible

Researchers successfully predict whether a bacterium is a plant pathogen

Researchers from UPM have developed a method that allows them to predict the infectivity of bacteria towards plants by using computational methods.

A research group from Centre for Plant Biotechnology and Genomics CBGP (UPM-INIA), a joint centre of Universidad Politécnica de Madrid (UPM) and Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA), has developed a computational tool able of automatically identifying known pathogenicity factors.

This online tool provides an instant catalog of the "weapons" available of the bacterium. From this data, the team of researchers created a predictor that is able to distinguish the genomes of pathogenic bacteria from the non-pathogenic bacteria with over 90% accuracy.

The emergence of new bacterial diseases associated to the plant consumption, such as the outbreak of Escherichia coli in Germany in 2011, makes it necessary a method that predicts the beginning of a bacterium from data of its genome. For better or for worse, bacteria dominate the world. But in general, for better.


E. coli bacteria. Credit: Wikimedia Commons

The majority of bacteria are innocuous or beneficial and without them it would be impossible life on Earth. However, a reduced number of bacterial species are able to infect other organisms and causing diseases. The plant pathogenic bacteria cause significant losses in crops, and what it is more, some human pathogens are related to plants at certain stages of their life cycle.

What does a bacterium need to infect another organism?  In the past, the microbiologists tried to answer this question by deactivating a particular bacterium gene and then measuring its infectivity. This strategy has been very effective, but it has limitations since it has allows microbiologists to identify a significant number of mechanisms but it fails to provide a global approach. Sequenced genomes of a large number of pathogenic and non-pathogenic bacteria have been available for a short time, and they have allowed researchers to tackle this question from a new approach: What genes must a bacterial genome contain to be able to infect a plant?

In order to give an answer to this question, a group of researchers from CBGP (UPM-INIA) has developed a tool called PIFAR (Plant-bacteria Interaction Factors Resource) that consists of a public repository of bacterial genetic determinants involved in the plant-bacterial interaction.

This tool can predict interactions between bacteria and plants, some interactions were already known but others were not. For instance, Cronobacter genus includes some opportunistic, foodborne pathogens that causes severe diseases in infants and newborns and can also colonizes plants. This model is able to identify plant-associated bacteria with a high precision, especially the Cronobacter turicensis strain that gives even a higher value and is particularly pernicious.

Pablo Rodríguez Palenzuela, a researcher of this study, says “the advances in genomics allow us to obtain the complete sequence of any bacterial species easily, even before the life cycle of the bacteria is studied”.

Therefore, this tool can predict whether a bacterium involved in a new outbreak has possibly a plant origin or not, this would help epidemiologists to control such outbreak”.

Martínez García, PM; López Solanilla, E; Ramos, C; Rodríguez Palenzuela, P. Prediction of bacterial associations with plants using a supervised machine-learning approach. ENVIRONMENTAL MICROBIOLOGY 18 (12): 4847-4861. DOI: 10.1111/1462-2920.13389. DEC 2016.

Fondo pie pagina
Politécnica Virtual
e-Adu
Ciuddad del futuro
Tienda UPM