During her time as a postdoctoral scientist at the University of Oxford, our Senior Lecturer Dr Clara Grazian worked on the largest ever global study of tuberculosis, which has identified genetic causes of drug resistance.
Using cutting-edge genomic sequencing techniques, the University of Oxford research team have identified almost all the genomic variation that gives people resistance to 13 of the most common tuberculosis (TB) drug treatments. The study's outcomes have just been released to the public.
Dr Grazian was in charge of the statistical analysis of the project between 2017 and 2019, working on the validation study first and on genome-wide association studies later to propose candidate mutations affecting tuberculosis resistance to antimicrobials. These were subsequently tested from a molecular point of view to better understand the specific biological mechanisms conferring resistance.
Her colleagues on the groundbreaking research study were from Oxford's Nuffield Department of Medicine and Big Data Institute.
“Statistical and machine learning techniques are central in analysing high-dimensional data including the whole genomes of ~20,000 strains of tuberculosis, and try to associate specific mutations to the development of resistance to tuberculosis”, Dr Clara Grazian said of her contribution to the study.
“In particular, advanced statistical techniques were used to reduce the biological noise of the data and to cluster strains depending on their response to antimicrobials, to identify few important mutation candidates. Moreover, machine learning techniques were used together with statistical techniques to increase the prediction performance of techniques based on the genome analysis, which can be used in place of expensive and time-consuming phenotype tests to understand the specific strain resistance patterns.”
For more on the study, see the press release.