Next-gen sequencing for clinical microbiology

Tuesday, December 21th, 2010 by hinrich

Structure of TMC207/R207910

Koen Andries pointed me recently to a publication by Pallen et al. in the October issue of Current Opinion in Microbiology. The article reviews the opportunities and challenges of using next-gen sequencing in the context of clinical microbiology. Applications such as studying pathogen biology and evolution, profiling of complex microbial communities, pathogen discovery, etc. are discussed.

In the context of translational research the authors mention our work on TMC207/R207910:

Whole-genome sequencing provides a rapid and convenient method for identifying the genetic determinants of resistance to antimicrobial agents. In a pioneering study, high-throughput sequencing of mycobacterial genomes pinpointed the mutations responsible for resistance to a new antimycobacterial diarylquinoline, thereby revealing the mode of action of the agent.

You can find the PubMed abstract and a link to the article here. Our original Science paper identifying the ATP synthase of Mycobacterium tuberculosis as the target of TMC207/R207910 can be found here.

Posted in Tuberculosis

Filtering on measurement reliability

Thursday, December 09th, 2010 by hinrich

PNAS article on I/NI calls

We have just published a comment on an article by Bourgon et al. in PNAS. The article was an initiative of Willem and highlights our positive experience with I/NI-calls.

In essence, next to filtering genes based on overall variance, we highlight the utility of filtering by reliability. What does that mean? If a given measurement can be repeated via different measurement devices (in our case: having multiple different probes on an Affymetrix microarray measuring the same transcript), we can assess the reliability of the measurement and use that to filter genes.

The idea is: the different measurement devices (the probes) should agree among one another when you look at the data across samples. If there is no agreement, then either the quality of some or all of your measurement devices is poor or there is no detectable biological difference between the samples beyond technical noise.

There are two conceptual advantages of this approach: we propose an approach to disentangle biological variation from technical noise and we are able to provide the user with a less arbitrary choice for a filtering threshold.

You can read the article here.

Posted in Molecular Profiling