The AI ​​tool reveals how TB drugs kill bacteria at the molecular level

Tuberculosis (TB) is the deadliest infectious disease in the world – and one of the most difficult to heal. Furthermore, Standard treatment requires a cocktail of several drugs over at least six months. However, and one in five patients has one type of tuberculosis that is withstanding these first -line drugs. For example, Now. Similarly, a new study offers a powerful AI assisted method to discover exactly how TB drugs kill bacteria, opening the door to smarter treatment combinations that could operate more quickly.

The development of more efficient and shorter treatment is a global priority. However, “We need a better multidrogue diet: three to five new drugs that work even for drugs currently resistant. However, ” explains Bree Aldridge, the main study of the study and professor of molecular biology and ai ​​tool reveals how tb microbiology at the University School of Medicine tuffs and professor in biomedical engineering at the Tofts University School of Engineering. For example, But progress has been slow. In addition, in part, because scientists lacked tools to see precisely how drugs work and therefore how they could work better together to attack TB bacteria.

“Tuberculosis probably has several Achilles heels that we could strike at the same time. Consequently, ” explains Aldridge, who is also an associate director of Stuart B. Consequently, Levy Center for Integrated Management of Antimicrobial Resistance at Tufts. Additionally, Consequently, “But it is surprisingly difficult to understand exactly how a drug kills its target cell. In addition, It is like entering a room. spy on bruised faces, a reversed chair and a broken lamp; You can say that a fight happened but not that started or how it went. In the same way. scientists can say when ai ​​tool reveals how tb a drug killed target cells but often not the exact chain of molecular events, AKA, “death mechanism”.

Aldridge. his collaborators of the TUFTS University School of Medicine and other institutions have now found a way to understand this mechanism. In a new study in Cellular systemsThey have shown how their new AI assisted tool – called Deciphaer (decoding intermodal information on pharmacologies via autoencoders) – can reveal. in molecular detail, how potential tuberculosis drugs kill bacteria.

The tool is based on the previous research of the team that has captured high resolution images of TB. bacteria at their death during treatment. These snapshots reveal clues – for example. changes in the form of bacterial cells or the internal structure – used by the attack mode of a drug. Scientists use this “morphological profiling” as a kind of crime scene investigation for cells: they dose TB bacteria with a new medication. ai ​​tool reveals how tb freeze them at the time of death and compare cellular damage resulting with the models observed from known antibiotics.

If you treat TB bacteria with a new medication. it goes in the same way as for other drugs that destroy the cell wall, you can assume that it also destroys the cell wall. “”

Bree Aldridge, principal author of the study

Using the AI, the team has now went further, connecting these visual indices to detailed readings of the activity of bacterial genes, called transcription profiles. The researchers have formed a model to identify molecular changes. such as bacterial genes that light up or deactivated, occur in parallel with specific visual changes.

“Before, we could only say about how a drug killed tuberculosis using morphological profiling. Now we can provide more accurate information on the impact of cells on cells and why bacteria die, ai ​​tool reveals how tb ”explains Aldridge. For example. by testing Deciphaer, she says that the team found that a clinical development TB drug did not work as expected.

“Based on similar existing compounds, we assumed that the drug worked by destroying the cell wall,” she says. “But he actually kills the TB bacteria by altering the respiratory chain and the capacity of cells to make energy. »»

Because the AI ​​tool can predict the molecular impact of a drug from single images – which is much cheaper than using RNA sequencing – it can reveal more quickly how potential TB treatments work in different growth conditions. genetic strains or drug combinations.

“We plan to continue using it in the combination of drugs in our own laboratory. hope that it will support collaborations around the world to accelerate the development of new TB drugs,” explains Aldridge. Although the need is particularly urgent for tuberculosis. she ai ​​tool reveals how tb adds that Deciphaer’s approach could also be applied to other infectious diseases and cancer.

William C. Johnson, a doctorate. Student in molecular microbiology with the Graduate School of Biomedical Science tuffs, is the first author. The research reported in this article was partly supported by the Gates Foundation. by the National Institutes of Health under the T32AI007422 reward number. Complete information on authors, donors, methodology, limitations and conflicts of interest is available in the published article. The content is only the responsibility of the authors and does not necessarily represent the official opinions of donors.