Imperial College London

First WHO catalogue of tuberculosis mutations developed with Imperial

by

3D computer image of tuberculosis

The largest catalogue of tuberculosis (TB) mutations has been developed by the WHO in collaboration with Imperial.


WHO catalgoue
Download the WHO catalogue

The catalogue for the Mycobacterium tuberculosis (Mtb) genome lists more than 17,000 mutations, and will help medics and health services around the world to interpret genome sequencing results and treat patients more quickly and using appropriate drug regimens.

Drug resistance to TB is becoming an increasing problem. Of the 10 million people estimated to have fallen ill with tuberculosis (TB) in 2019, nearly half a million developed TB resistant to the antibiotic rifampicin (RIF).

The majority of patients with RIF-resistant TB are able to be detected with existing tests, but the situation is not that clear for other anti-TB drugs due to the limited knowledge of how resistance is associated with mutations in the Mycobacterium tuberculosis (Mtb) genome. 

“Tuberculosis is the largest infectious killer in the world today and drug resistance is becoming a significant global problem." Dr Leonid Chindelevitch School of Public Health

The report, based on analysing the largest globally diverse dataset of Mtb genomes so far, lists mutations, their frequency and the degree of their association with resistance and includes methods used, mutations identified and summaries of important findings for each drug.

Researcher Dr Leonid Chindelevitch, from the School of Public Health, said: “Tuberculosis is the largest infectious killer in the world today and drug resistance is becoming a significant global problem.

"The WHO catalogue will help medics and health services around the world to interpret results and provide faster and more targeted treatment for patients of this deadly disease.”

The TB mutations catalogue work was supported by FIND and Unitaid.

Interpreting complex strains

Dr Chindelevitch has also developed a computer programme to determine the individual strains of TB in patients that are infected with a complex infection consisting of multiple strains.

The research, recently published in Microbial Genomics, uses a statistical model to ‘disentangle’ the underlying strains of TB from genome sequencing data.

Dr Chindelevitch says that the tool could speed up the process of diagnosis and treatment.

Currently if a medic suspects a patient has TB, a sample is sent to a laboratory for sequencing and the bacteria is grown to determine drug sensitivity to the underlying strains.

This process can typically take two to three days, causing critical delays to appropriate patient treatment. This tool could enable much faster determination of the strains, leading to the patient receiving the appropriate treatment within hours.

Identifying variants

Further research, published in Algorithms for Molecular Biology, demonstrates how machine learning can be used to identify additional mutations in TB that may make it resistant to some antibiotics.

The researchers expect the INGOT-DR method to become a key part of the drug resistance prediction toolkit for clinical and public health microbiology researchers.

Study author Dr Hooman Zabeti, from Simon Fraser University in Canada and a visiting researcher at Imperial, said: "Using an interpretable but flexible method such as INGOT-DR can provide both the ability to accurately predict drug resistance in TB as well as identify putative resistance variants that can be independently verified.

"It is remarkable that, when given no prior information other than a large dataset, INGOT-DR identified many variants known to be involved in drug resistance."

---

The research was supported by the Medical Research Council and Genome Canada.

Supporters

Reporter

Stephen Johns

Stephen Johns
School of Public Health

Click to expand or contract

Contact details

Tel: +44 (0)20 7594 9531
Email: s.johns@imperial.ac.uk

Show all stories by this author

Tags:

International, Global-challenges-Health-and-wellbeing, Tuberculosis, Global-health
See more tags