The current method for assessing antibiotic resistance requires cultivation of the accused bacteria (s). It requires at least 2 days. In an emergency, this period may be far too long. An Inria (National Institute for Research in Digital Sciences and Technology Institute) is in the process of developing a technique allowing it to shorten it for a few hours, even a few minutes.
It starts from the principle that antibiotic resistance is accompanied by changes in the bacterial genome. Logically, this observation implies the sequencing of this genome in search of changes in question. But then again, the deadlines for obtaining a result are long. At the Harvard (United States) Medicine School, the bioinformatics researcher Karel Břinda has initiated a new technique that he is currently developing within the Inria Center at the University of Rennes. It consists in comparing the bacterial DNA fragments obtained by sequencing with those of a bacterial DNA data bank corresponding to bacteria known to be sensitive or resistant to antibiotics. The starting hypothesis is that the more the DNA sequences of the tested bacteria are close to those of antibiotic resistant bacteria, the higher the chances that this bacteria is also resistant.
A sequencing device not larger than a smartphone
This so -called genomic neighbor’s typing method (Genomic Neighbor Typing) At 2 advantages. First, it does not require any information on the complex, complex functioning of the bacteria tested. Second, the result can be obtained from a very small amount of data. Thus, explains Karel Břinda, “ We have shown that once the sequencing started, we are able to predict resistance and sensitivity in approximately 10 minutes ».
This technique is facilitated by the use of so -called ” nanopores », No more bulky than a smartphone. With them, continues Karel Břinda, ” You receive a data flow as soon as the device begins to sequence. And our method makes it possible to carry out a prediction almost immediately from these first data. The diagnosis therefore arrives very quickly. In addition, the sequences obtained being quite large, ” As soon as you have a few long enough pieces, you can make a good estimate of the pathogen resistance. »
The essential clinical expertise
The technique is therefore promising, but it comes up against a major obstacle: it has been validated with pneumococcus and gonococcus, but for the moment, it cannot be used for all antibiotics and all bacteria. The reason is simple: the databases are not complete. Consequently, the INRIA team has teamed up with the CHU (University Hospital Center) of Rennes for “ Connect your computational expertise to the biological and clinical skills of the CHU to walk to methods that are directly applicable in the hospital context. The CHU indeed has a team that manages the collections of the CNR (National Reference Center for Antibiotic Resistance) for enterococci. More generally, explains Karel Břinda, ” The key question ” East ” The construction of databases that are both large and representative for bacterial strains. “Because of technical progress, these bases are called upon to develop considerably, which implies the development of” new computational methods and new software tools ».
