Abstracts

Tuberculosis, caused by bacteria in the Mycobacterium tuberculosis complex, remains a major global health challenge, with drug resistance and the absence of an effective vaccine continuing to hinder control efforts. Advances in sequencing technologies and analytical methods — including bioinformatics, data science, and AI — are transforming the generation and interpretation of ’omics data, enabling the design of improved diagnostics, therapies, and vaccines.

Many countries now use whole-genome and amplicon-based sequencing to characterise circulating M. tuberculosis strains, uncover transmission clusters, and detect genotypic drug-resistance variants, providing actionable intelligence for clinical settings, surveillance, and infection-control programmes. Evidence of host–pathogen co-evolution is also driving genome-to-genome studies that seek to pinpoint genetic interactions shaping disease risk, progression, and treatment outcomes.

This talk will highlight applications of ’omics — particularly genomics and transcriptomics — across both host and pathogen, and will discuss emerging opportunities, including the use of AI-driven approaches, to accelerate insights and deliver tools that support global tuberculosis control.

During this presentation we tackle the issue of the global and local history of tuberculosis through space and time. We provide an up-to-date synthesis on evolutionary issues related to tuberculosis origin, ecology, and current academic research, focusing on recent results obtained using comparative genomics of Mycobacterium tuberculosis complex (MTBC) infections.

Key questions include: How sure are we that MTBC was a human disease at emergence? How old is the most recent common ancestor between animal and human MTBC ecotypes? What is the relative importance of human demography and bacterial transmissibility in the epidemic success of tuberculosis? What kind of host–pathogen adaptation are we dealing with: co-divergence or co-evolution?

After reviewing major paradigm shifts arising from genomic research, this talk highlights the diversity of unresolved questions and shows how understanding the past of the pandemic may help guide future strategies to eradicate the disease.

Tuberculosis (TB) control in high-burden settings such as Thailand requires both precise genomic surveillance of drug-resistant strains and innovative approaches for population-level screening. Multidrug-resistant (MDR) and extensively drug-resistant (XDR) TB continue to spread, while latent TB infection (LTBI) remains an underdiagnosed reservoir for future disease.

To address these challenges, we combined four complementary approaches. First, whole-genome sequencing (WGS) of MDR/XDR isolates from hospitals and provincial laboratories nationwide revealed multiple interprovincial transmission clusters and silent spread of XDR strains, underscoring the urgent need for cross-province genomic surveillance.

Second, we systematically evaluated and updated the drug-resistance (DR) mutation catalog, identifying rare variants in second-line drug targets that were absent from existing references but correlated with phenotypic resistance. This improved the accuracy of mutation-based diagnostics and provided insights into emerging resistance mechanisms.

Third, we developed TBLandScape, a national genomic analysis platform incorporating 3,354 Mycobacterium tuberculosis genomes to establish Thailand’s first comprehensive TB genomic database. The platform integrates drug-resistance prediction, lineage classification, and phylogenetic analysis through a user-friendly interface equipped with customizable mutation databases, IGV visualization, and geo-temporal analytics (Phylomap). It supports both short- and long-read sequencing data, facilitating applications in clinical, surveillance, and research contexts. Validation using 594 reference samples demonstrated >98% accuracy for lineage and resistance prediction compared with TB-Profiler, confirming its reliability as a national bioinformatics resource.

Finally, to address the lack of effective LTBI screening tools, we evaluated a label-free surface-enhanced Raman spectroscopy (SERS) approach using 1,000 plasma samples from Northeast Thailand, equally divided between IGRA-positive and negative individuals. Raman mapping (7 × 7 grid) was completed within 10 minutes per sample. Optimized machine-learning models achieved 81% accuracy in train-test analysis and 75% in leave-one-out cross-validation across all batches, improving to 93% with optimized chip design and preprocessing using logistic regression.

Collectively, the integration of genomics, informatics, and AI-enhanced spectroscopy provides a rapid, scalable framework to strengthen TB diagnosis, surveillance, and LTBI screening — accelerating Thailand’s progress toward the End TB Goal.

Mycobacterium tuberculosis lineage 2 (L2) remains a globally significant lineage associated with increased drug resistance and rapid transmission. The L2 lineage exhibits a hotspot for genetic diversity and evolution in Panama, requiring in-depth analysis.

In this talk, we present a prospective analysis of Mycobacterium tuberculosis L2 isolates from Colon City, Panama. Using ASO-PCR, we identified 31.7% (86/271) of isolates as Modern L2.2. Whole-genome sequencing confirmed all isolates belonged to the L2.2.1 sublineage, with 96.9% (62/64) classified as pan-susceptible and 3.1% (2/64) as rifampicin/pyrazinamide-resistant.

Sublineage analysis based on SNPs using the TB-gen tool identified a mutation at position 1219683G > A, genotyping all strains as the L2.2.M3 sublineage. A correlation with geographical distribution was observed compared with other Latin American L2 isolates, alongside a relatively low evolutionary rate within Panama.

Using the TB annotator tool to compare with 1,578 L2.2.M3 strains worldwide, the Panamanian strains formed a distinct monophyletic cluster within the sublineage. This cluster contained all strains from Colon Province and three additional strains identified in non-Asian countries. The closest related branches had higher proportions of drug-resistant strains, primarily from East and Southeast Asia.

Together, these findings suggest endemic transmission of the Mycobacterium tuberculosis L2.2.M3 sublineage in Colon, Panama. We recommend combining genomic information with epidemiological data to accurately track and identify transmission hotspots locally and globally.

Full author list: Igor Mokrousov (1), Tatiana Vinogradova (2), Natalia Solovieva (2), Ivaylo Slavchev (3), Marine Dogonadze (2), Dmitry Polev (1), Georgi Dobrikov (3), Violina T. Angelova (4), Violeta Valcheva (5), and Anna Vyazovaya (1)

(1) St. Petersburg Pasteur Institute, St. Petersburg, Russia
(2) St. Petersburg Research Institute of Phthisiopulmonology, St. Petersburg, Russia
(3) Institute of Organic Chemistry with Center of Phytochemistry, Bulgarian Academy of Sciences, Sofia, Bulgaria
(4) Medical University of Sofia, Sofia, Bulgaria
(5) Stephan Angeloff Institute of Microbiology, Bulgarian Academy of Sciences, Sofia, Bulgaria

Different spontaneous mutations emerge in Mycobacterium tuberculosis, and those beneficial for bacterial survival are selected, leading to drug resistance, tolerance, and persistence. We studied genetic variation in response to selective pressure from antibiotics and candidate anti-tuberculosis compounds in both in vivo and in vitro models.

C57Bl/6 mice were infected with different MDR clinical strains and treated with moxifloxacin, linezolid, and bedaquiline. Bacterial isolates were recovered from lungs after 2 and 5.5 months of treatment. For the in vitro study, the H37Rv reference strain was cultured under elevated concentrations of new candidate anti-TB compounds (nitrofuranes and aroylhydrazones), and resistant clones were subjected to whole-genome sequencing.

Treatment reduced bacterial burden in mouse lungs, and no resistance mutations to new drugs emerged. However, some isolates showed mutations beyond drug resistance. In vitro studies revealed responses to nitrofuranyl amide via multiple pathways counteracting oxidative and nitrosative stress. Mutations were also detected in genes linked to drug tolerance and efflux mechanisms.

Long-term treatment of mice infected with a hypervirulent strain resulted in selection of mycobacteria carrying a mutation inactivating the tgs3 gene, associated with lipid metabolism and dormancy. Its inactivation led to increased bacterial growth. The in vitro study highlighted complex responses to nitrofuranes and aroylhydrazones, with mutations emerging in nonspecific tolerance mechanisms and stress-response pathways.

This study was supported by the Russian Science Foundation (grant 24-44-00004).

Full author list: Danila Zimenkov¹, Anastasia Ushtanit¹, Marina Filippova¹, Uliana Semenova¹, Vyacheslav Zhuravlev², Natalia Solovieva², Peter Yablonsky², Maria Sviridenko³, Anastasia Khakhalina³, Svetlana Safonova³, Marina Makarova³, Elizaveta Gordeeva⁴, Elena Guselnikova⁴, Yakov Schwartz⁴, Natalia Stavitskaya⁴, Yuliana Atanasova⁵, Stanislava Yordanova⁵, Ana Baykova⁵, Elizabeta Bachiyska⁵, Igor Mokrousov⁶

¹ Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, Moscow, Russia
² Saint-Petersburg State Research Institute of Phthisiopulmonology, Russia
³ Moscow Research and Clinical Center for Tuberculosis Control, Russia
⁴ Novosibirsk TB Research Institute, Russia
⁵ National Reference Laboratory of Tuberculosis, National Center of Infectious and Parasitic Diseases, Sofia, Bulgaria
⁶ St. Petersburg Pasteur Institute, St. Petersburg, Russia

Infections caused by nontuberculous mycobacteria are becoming increasingly significant due to the growing number of vulnerable individuals worldwide. Understanding evolutionary relationships within the genus Mycobacterium is critical for improving species identification and enhancing diagnosis, treatment, and epidemiological tracking.

Comparative genomic analyses using average nucleotide identity, genome–genome distance, Mash values, multilocus sequence analysis, and average amino acid identity (AAI) demonstrated that AAI is the most effective metric for distinguishing Mycobacterium from other genera of Mycobacteriales, producing phylogenetic trees with minimal topology error.

Genes encoding 16S and 23S rRNAs also supported genus delineation. The established 94.5–95.0% identity threshold for the rrs gene was confirmed, while an analogous threshold of 88.5–89.0% was estimated for the rrl gene.

These findings do not support the proposed division of Mycobacterium into five genera. However, AAI distance distribution suggests the potential existence of a separate genus corresponding to the M. chelonae–abscessus complex, though this remains uncertain and requires clinical considerations. Overall, at least 402 distinct mycobacterial species were identified, 246 of which have been detected in clinical human specimens.

A gyrB fragment–based hybridization assay (“Myco-biochip”) was developed and evaluated in clinical antituberculosis centres in Moscow, Saint Petersburg, Novosibirsk, and Sofia between 2022 and 2024. In total, 71 mycobacterial species were identified across 3,119 samples from 2,221 patients in Russia and 48 samples from 48 patients in Bulgaria. Laboratory testing confirmed reliable identification of more than 80 species.

Four novel mycobacterial species related to M. duvalii, M. lentiflavum, M. talmoniae, and M. iranicum were identified across multiple sites. The discovery of a close relative of M. talmoniae supports the existence of a distinct clade positioned between M. terrae, M. triviale, and other slow-growing mycobacteria, further arguing against splitting the genus.

Acid-fast bacilli identified in tuberculosis-suspected patients were not limited to Mycobacterium, but also included species from Nocardia, Gordonia, Corynebacterium, Tsukamurella, and Rhodococcus within the order Mycobacteriales. These findings highlight the importance of accurate species identification and genotyping for epidemiology, public health strategy, and improved diagnostic and treatment approaches.

Full author list: Ratchanon Sophonmanee¹, Elena Stylianou², Helen McShane², Nawamin Pinpathomrat¹*

¹ Department of Biomedical Sciences and Biomedical Engineering, Faculty of Medicine, Prince of Songkla University, Hat Yai, Songkhla, Thailand
² The Jenner Institute, University of Oxford, Old Road Campus Research Building, Roosevelt Drive, Oxford, United Kingdom

Tuberculosis (TB) remains one of the world’s leading infectious causes of mortality, underscoring the urgent need for more effective and durable vaccines beyond the century-old Bacille Calmette–Guérin (BCG). Advances in immunology, molecular biology, and biomaterials have accelerated the development of next-generation TB vaccine platforms designed to induce robust, long-lasting, and tissue-specific immunity.

This talk provides an overview of emerging TB vaccine platforms, including live-attenuated and recombinant mycobacterial vaccines, viral-vectored vaccines, protein subunit vaccines with novel adjuvants, and nucleic acid–based approaches such as mRNA vaccines.

Equally critical to vaccine efficacy is the delivery system, which governs antigen presentation, immune polarization, dose-sparing potential, and scalability. The presentation will explore innovative delivery strategies including lipid nanoparticle systems, viral and non-viral vectors, and skin-targeted delivery methods such as microneedle arrays that leverage the high density of antigen-presenting cells in the dermis.

Special emphasis will be placed on tailoring delivery platforms to enhance cellular immunity, particularly Th1 and CD8⁺ T-cell responses, which are central to protection against Mycobacterium tuberculosis.

By integrating advances in vaccine platform design with optimized delivery technologies, this talk highlights a translational pathway toward more effective TB vaccination strategies. Key challenges in clinical development, manufacturing, and implementation in high-burden settings will also be discussed, outlining future directions for TB vaccine innovation in the era of precision vaccinology.

This study aimed to clarify tuberculosis (TB) transmission patterns — including recent transmission versus endogenous reactivation — as well as identify high-risk groups and transmission hotspots in urban and rural China using genomic epidemiological methods. Whole-genome sequencing combined with epidemiological investigation was applied to generate evidence for optimizing TB control strategies in high-burden settings.

Conducted from 2009 to 2023 across urban sites in Shanghai Songjiang and Shenzhen Longhua, and rural sites in Sichuan Wusheng, Heilongjiang Wuchang, and Henan Linzhou, the study defined recent transmission as ≤12 single nucleotide polymorphisms (SNPs). A total of 2,212 urban and 2,418 rural culture-positive pulmonary TB patients were analyzed.

Findings revealed marked urban–rural disparities in TB transmission. Urban areas showed relatively low recent transmission clustering rates (25.2% in Songjiang and 15.4% in Longhua), with 70–74.7% of patients being migrants. Transmission occurred mainly within migrant populations (61%) and local residents (17%), with unmanaged migrants identified as major sources of infection. Approximately 70% of migrants developed TB within two years of arriving in cities, and transmission risk declined significantly with increasing spatial distance.

Rural areas demonstrated higher clustering rates (26.9–44.7%), with Mycobacterium tuberculosis predominantly belonging to lineage L2 (72.8%). Close contacts accounted for 42% of cases, while students showed a 3.84–4.82-fold higher clustering risk. Transmission patterns also varied across geographical regions.

The study indicates that TB transmission in high-burden areas is largely driven by recent transmission, although clustering may be underestimated due to sequencing limitations and population mobility. Models focused solely on high-risk groups from developed countries may not be directly applicable to China.

Active case finding remains central to TB control, requiring context-specific identification of high-risk groups and hotspots. Strengthening migrant health management, student protection, and monitoring of high-density gathering settings will be essential to reduce community transmission and improve TB control outcomes.

We have recently conducted several human whole-genome sequencing (WGS) projects and investigated human genomic variation associated with a wide range of diseases, including infectious diseases. Genome-wide association studies (GWAS) and high-resolution typing of human leukocyte antigen (HLA) genes were performed to identify genetic factors influencing disease susceptibility, drug response, and vaccine response.

One notable finding involves interactions between pathogen genomes and the human genome. Interactive effects between hepatitis B virus (HBV) genomic mutations and human HLA genes were identified, where specific combinations of HBV variants and HLA-DPB1 alleles were associated with increased risk of progression from chronic hepatitis to hepatocellular carcinoma.

Both high and low responses to HB vaccination were significantly associated with specific HLA-DRB1 alleles. Associations were also observed between adverse effects following COVID-19 vaccination and human genetic variation, including variants in the IL1RL1/IL18R1/IL18RAP region and in HLA genes.

We also present findings on interactions between tuberculosis lineages and human genomic variation. In addition, we contribute to the “Infectious Disease Clinical Research Network with National Repository” by performing whole-genome sequencing of patient samples.

While global TB control strategies have historically focused on the genome of Mycobacterium tuberculosis, the genetic architecture of the human host remains an overlooked factor in determining clinical outcomes. Achieving End TB targets requires integrating host genetic profiling alongside pathogen surveillance.

This presentation highlights the urgent implementation of pharmacogenomics (PGx) to mitigate drug-induced liver injury (DILI), a major cause of treatment interruption and mortality. Particular focus is placed on NAT2 (N-acetyltransferase 2), where slow acetylator alleles common in Southeast Asian populations can lead to toxic accumulation during standard isoniazid dosing. The role of rapid acetylators in increased mortality, particularly among HIV-infected TB patients, is also explored.

Beyond treatment toxicity, host susceptibility is examined through pathogen–host co-evolution. Lineage-specific genetic risk is highlighted, including associations between HLA-DRB1*09 and susceptibility to Lineage 2 (Beijing family) strains, and CD53 variants linked to Lineage 1 (Indo-Oceanic) infection. These findings suggest that host immune responses are often dependent on the infecting strain.

The study integrates host susceptibility data with whole-genome sequencing (WGS) of Mycobacterium tuberculosis to map transmission dynamics. WGS enables detection of cryptic outbreak clusters and differentiation between recent transmission and reactivation. Overlaying host genetic risk onto WGS-defined transmission networks helps identify high-risk transmission nodes where specific host–pathogen combinations accelerate community spread.

A “Dual-Genome Triage” model is proposed, combining regional NAT2 genotyping with pathogen WGS to stratify patients into standard versus precision dosing pathways while simultaneously disrupting transmission chains. This host-informed, pathogen-aware approach represents a key step toward reducing preventable deaths and advancing TB elimination efforts.

Keywords: Host Genetics, NAT2, Pharmacogenomics, HLA-DRB1, CD53, Whole Genome Sequencing (WGS)