Hoping to spur better diagnostic tools and new therapies for drug-resistant tuberculosis, HSPH researchers and their colleagues have put together a comprehensive open-access database of drug resistance mutations.
Increasing resistance to the mainstay drugs isoniazid and rifampicin and also to many of the second-line options threatens global efforts to control TB. Diagnosis everywhere now depends on culturing the slow-growing organism, which is nearly impossible in resource-poor settings, said senior author Megan Murray, HSPH associate professor of epidemiology.
“It takes so long that by the time people are diagnosed, they have died or transmitted it to others, or both,” she said. “Part of the goal is to develop cheap and rapid diagnostics.”
The TB Drug Resistance Mutation Database (http://www.tbdreamdb.com) is divided into two parts, said creator and curator Andreas Sandgren, a postdoctoral fellow. The first part is a collection of unique mutations associated with resistance to specific drugs culled from a systematic review of hundreds of scientific papers.
For the second part, Sandgren extracted the relative frequency of the most common mutations from 10 high-quality publications, which is helpful for geographic surveillance and diagnostic strategies, he said.
The database started as a tool to support two major projects in the Murray group, the first a whole-genome sequencing of three dozen drug-resistant TB strains from patients around the world and the second an in-depth analysis of putative drug-resistance genes in several thousand clinical samples. This work is in progress. Published results will be added to the database.
“After we created it, we realized it would be useful for the rest of the TB research community,” said Sandgren, first author of the paper announcing the resource in the February PLoS Medicine.
“So many people need to use these data,” said Murray. “To waste effort and time to recompile them again and again is crazy. This can really change the way research occurs. It means people can bring their own ideas to research the database without having to do the actual work of compiling it.”
As of September, the database contained 946 unique mutations associated with seven different drug classes and spread over 36 genes, two intergenic/promoter regions, and one ribosomal RNA coding region. “We set up strict criteria to make sure we can trust the data,” said Sandgren, who will add data as newly published research warrants.
Sandgren will continue to curate the database and collaborate on the project when he returns to the Karolinska Institute to continue his postdoctoral studies.
Students may contact Megan Murray at mmurray@hsph.harvard.edu for more information.
Conflict Disclosure: The authors declare no conflicts of interest.
Funding Sources: The Ellison Foundation. Sandgren was also funded through a postdoctoral stipend from the Swedish Society for Medical Research/Stiftelsen för Medicinsk Forskning and Stiftelsen Olle Engqvist Byggmästare; the content of the work is the responsibility solely of the authors
