A new method for detecting mixed Mycobacterium tuberculosis infection and reconstructing constituent strains provides insights into transmission
Bibliographic record
Abstract
BACKGROUND: Mixed infection with multiple strains of the same pathogen in a single host can present clinical and analytical challenges. Whole genome sequence (WGS) data can identify signals of multiple strains in samples, though the precision of previous methods can be improved. Here, we present MixInfect2, a new tool to accurately detect mixed samples from Mycobacterium tuberculosis short-read WGS data. We then evaluate three approaches for reconstructing the underlying mixed constituent strain sequences. This allows these samples to be included in downstream analysis to gain insights into the epidemiology and transmission of mixed infections. METHODS: We employed a Gaussian mixture model to cluster allele frequencies at mixed sites (hSNPs) in each sample to identify signals of multiple strains. Building upon our previous tool, MixInfect, we increased the accuracy of classifying in vitro mixed samples through multiple improvements to the bioinformatic pipeline. Major and minor proportion constituent strains were reconstructed using three approaches and assessed by comparing the estimated sequence to the known constituent strain sequence. Lastly, mixed infections in a real-world Mycobacterium tuberculosis population from Moldova were detected with MixInfect2 and clusters of recent transmission that included major and minor constituent strains were built. RESULTS: All 36/36 in vitro mixed and 12/12 non-mixed samples were correctly classified with MixInfect2, and major strain proportions were estimated with high accuracy (within 3% of the true strain proportion), outperforming previous tools. Reconstructed major strain sequences closely matched the true constituent sequence by taking the allele at the highest frequency at hSNPs, while the best-performing approach to reconstruct the minor proportion strain sequence was identifying the closest non-mixed isolate in the same population, though no approach was effective when the minor strain proportion was at 5%. Finally, fewer mixed infections were identified in Moldova than previous estimates (6.6% vs 17.4%) and we found multiple instances where the constituent strains of mixed samples were present in transmission clusters. CONCLUSIONS: MixInfect2 accurately detects samples with evidence of mixed infection from short-read WGS data and provides an excellent estimate of the mixture proportions. While there are limitations in reconstructing the constituent strain sequences of mixed samples, we present recommendations for the best approach to include these isolates in further analyses.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".