Rapid, accurate, and reproducible de novo prediction of resistance to antituberculars

Abstract

As one of the deadliest infectious diseases in the world, tuberculosis is responsible for millions of new cases and deaths reported annually. The rise of drug-resistant tuberculosis, particularly resistance to first-line treatments like rifampicin, presents a critical challenge for global health, which complicates the treatment strategies and calls for effective diagnostic and predictive tools. In this study, we apply an ensemble-based molecular dynamics computer simulation method, TIES_PM, to estimate the binding affinity through free energy calculations and predict rifampicin resistance in RNA polymerase. By analyzing 61 mutations, including those in the rifampicin resistance-determining region, TIES_PM produces reliable results in good agreement with clinical reference and identifies abnormal data points indicating alternative mechanisms of resistance. In the future, TIES_PM is capable of identifying and selecting leads with a lower risk of resistance evolution and, for smaller proteins, it may systematically predict antibiotic resistance by analyzing all possible codon permutations. Moreover, its flexibility allows for extending predictions to other first-line drugs and drug-resistant diseases. TIES_PM provides a rapid, accurate, low-cost, and scalable supplement to current diagnostic pipelines, particularly for drug resistance screening in both research and clinical domains.IMPORTANCEAntimicrobial resistance (AMR), a global threat, challenges early diagnosis and treatment of tuberculosis (TB). This study employs TIES_PM, a free-energy calculation method, to efficiently predict AMR by quantifying how mutations in bacterial RNA polymerase (RNAP) affect rifampicin (RIF) binding. On simulating 61 clinically observed mutations, the results align with WHO classifications and reveal ambiguous cases, suggesting alternative resistance mechanisms. Each mutation requires ~5 h, offering rapid, cost-effective predictions. An ensemble approach ensures statistical robustness. TIES_PM can be extended to smaller proteins for systematic codon permutation analysis, enabling comprehensive antibiotic resistance prediction, or adapted to identify low-resistance-risk drug leads. It also applies to other TB drugs and resistant pathogens, supporting personalized therapy and global AMR surveillance. This work provides novel tools to refine resistance mutation databases and phenotypic classification standards, enhancing early diagnosis while advancing translational research and infectious disease control.

Keywords: computational biology; drug resistance prediction; rifampicin resistance; tuberculosis.

Fig 1

Changes in Gibbs free energy (ΔΔG) in protein-drug binding with and without mutation. State A represents the bound complex (protein + drug), and state B represents the unbound state (protein + drug). The presence of a mutation introduces states A′ (mutant complex) and B′ (mutant protein + drug). ${\displaystyle \mathrm{\Delta }{\mathrm{G}}_{\mathrm{c}\mathrm{o}\mathrm{m}}^{}}$ and ${\displaystyle \mathrm{\Delta }{\mathrm{G}}_{\mathrm{p}\mathrm{r}\mathrm{o}\mathrm{t}}^{}}$ denote the changes in free energy for the respective states, whereas ${\displaystyle \mathrm{\Delta }{\mathrm{G}}_{\mathrm{b}\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{i}\mathrm{n}\mathrm{g}}^{\mathrm{w}\mathrm{t}}}$ and ${\displaystyle \mathrm{\Delta }{\mathrm{G}}_{\mathrm{b}\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{i}\mathrm{n}\mathrm{g}}^{\mathrm{m}\mathrm{u}\mathrm{t}}}$ reflect the binding free energy differences before and after mutation.

Fig 2

The alchemical transformation process and ensemble simulation setup. Figure (A) illustrates the relationship between the degree of the mutation parameter λ and the alchemical mixture states. 0 < λ < 1 here are the intermediate transformation states, and although three states are shown here, in practice, this interval includes a number of states. Here, in the current work, eleven intermediate λ states are utilized along with the two ends (λ = 0 and λ = 1), giving a total of 13 states. Figure (B) shows the ensemble simulation architecture, where five replicas are employed for every state (41).

Fig 3

M. tuberculosis RNA polymerase (RNAP) structures. (A) shows the relative position of rpoB (cyan) in the protein, and all RNAP subunits are shown in surface view, with nucleic acids omitted in close-ups; (B) shows the mutations within rpoB, depicted as associated with antibiotic resistance (red), susceptibility (blue), and unknown type (gray), positioned relative to antibiotic binding sites (yellow); and (C) shows the spatial distribution of mutations relative to RRDR region and the drug-binding site, grouped by clinical phenotypes: R (resistant), S (susceptible), U (unknown within RRDR), and U (unknown outside RRDR).

Fig 4

Our results align well with both clinical data and the previous study (31). The impact of the listed mutations on the free energy of rifampicin binding to RNAP is presented. We display the deep-colored bars exhibiting our findings, and the light-colored bars exhibit those of the previous investigation (31). Susceptible (S) mutations, as revealed by clinical screening, are colored blue, whereas resistant (R) mutations are in red. The dashed line marks the ΔΔG threshold corresponding to the epidemiological cutoff for rifampicin resistance; mutations with values above this threshold are considered resistant in M. tuberculosis. Error bars represent 95% confidence intervals.

Fig 5

The calculated impact of 61 mutations on rifampicin’s binding free energy to RNAP. Mutations are categorized into two groups based on WHO clinical classifications: (A) mutations classified as resistant or susceptible in the WHO data set, represented in red and blue, respectively; (B) mutations initially classified as unknown due to insufficient or conflicting clinical evidence. In (B), all mutations are classified as unknown, but those occurring in the RRDR region are shown in light red and conservatively classified as resistant by WHO for safety considerations. Dashed lines mark the ΔΔG value corresponding to the epidemiological cutoff for rifampicin; values above this indicate clinical resistance in M. tuberculosis. Bars represent the mean ΔΔG for each mutation compared to the wild type, with 95% confidence intervals displayed.

Fig 6

Potential factors for susceptibility predicted in three mutations. (A) shows the structure changes in I491F, where the benzene ring of Phenylalanine and the benzene ring in RIF end up parallel to each other. (B) and (C) show the structure changes in L452P and H445C.