PREreview of Integrating measles wastewater and clinical whole-genome sequencing enables high-resolution tracking of virus evolution and transmission

Summary

This preprint presents an integrated wastewater and clinical whole-genome sequencing approach for measles virus surveillance during the 2024-2025 outbreak in South Africa. The authors analyzed 4,502 wastewater samples, identified measles RNA in 395, sequenced a subset, and combined these data with clinical whole-genome sequences to track genotypes B3 and D8. The study shows that wastewater measles positivity broadly reflected clinical case patterns, that wastewater sequencing recovered genotype shifts over time, and that whole-genome sequencing provided substantially higher resolution than standard N450 sequencing for reconstructing geographic spread and transmission links.

This work moves the field forward by demonstrating that wastewater-based genomic surveillance can complement case-based measles surveillance, especially where clinical sampling is uneven or sequencing coverage is limited. The manuscript is timely and useful because measles surveillance needs scalable tools that can detect transmission at population level and help distinguish local spread from multiple introductions.

Positive Feedback

The study is strong in its integration of laboratory, epidemiological, and phylogenetic approaches. I especially appreciated the comparison of wastewater and clinical surveillance, the validation of Freyja-based genotype assignment, and the direct comparison between whole-genome and N450-based analyses. The public health motivation is clear, and the work provides a practical model for expanding genomic surveillance beyond SARS-CoV-2 into vaccine-preventable diseases.

Major Issues

1. The representativeness of wastewater sequencing should be clarified. Although the study is framed as national surveillance, most wastewater sequence data appear to come from Gauteng. The authors should provide a clearer breakdown of sequenced and successfully genotyped wastewater samples by province, district, site type, and time period. This would help readers assess how strongly the conclusions apply across South Africa versus mainly to Gauteng.

2. Only a subset of positive wastewater samples was sequenced, and only 51/395 positives had sufficient coverage for genotype prevalence estimates. The manuscript would be strengthened by comparing viral load, site, date, and geography between sequenced/analyzable and non-analyzable positives. This would make it easier to evaluate whether genotype prevalence estimates could be biased toward higher viral-load samples or better-covered sites.

3. The transmission-inference language could be more cautious. The phylogenetic results are compelling, but phrases implying directionality, such as wastewater sequences descending from clinical sequences or evidence of interprovincial spread, should be supported with clearer uncertainty estimates. Please consider adding bootstrap/support values, temporal-signal diagnostics, sensitivity analyses using different coverage thresholds, and a clearer discussion of how incomplete clinical and global background sampling may affect inferred transmission links.

4. The comparison between whole-genome sequencing and N450 sequencing is important, but the fairness of the comparison should be described more explicitly. For example, were the same sequences, filters, and phylogenetic methods used after masking to N450? How were sequences without N450 coverage handled? The AU and parsimony-score analyses support the authors’ conclusion, but a short explanation of their interpretation for non-phylogenetics readers would improve clarity.

5. The wastewater-clinical correlation analyses could be expanded. District-level wastewater TPR correlated strongly with case counts, while temporal clinical and wastewater measures were weaker or not correlated depending on the clinical metric. Lagged analyses, adjustment for testing volume, population/catchment size, and the wastewater concentration-method change in week 41 of 2024 would help clarify how wastewater signals should be interpreted operationally.

Minor Issues

1. There is a date inconsistency: the methods and results describe samples through August 31, 2025, but some figure captions refer to October 1, 2024 to September 30, 2025. Please check and harmonize these dates.

2. Please define abbreviations clearly at first use, especially WES, WWTP, TPR, SNV, CMH, PS, and AU.

3. The clinical whole-genome sequencing subset of 39 samples should be described in more detail. Were these selected by geography, viral load, storage availability, genotype, or epidemiological relevance?

4. The change from concentration method C to method A is important. A brief statement on whether this affected longitudinal comparability of wastewater viral load or positivity would help.

5. The figures contain a lot of information. Adding clearer site labels, catchment/population context, and color-blind-accessible genotype colors would improve readability.

6. The data and code availability statements are helpful. Please include a specific repository release, commit hash, or analysis environment so the results can be reproduced more easily.

7. The discussion would benefit from a short dedicated limitations paragraph covering Gauteng-heavy sequencing, incomplete clinical sequencing, low-viral-load wastewater failures, and uncertainty in phylogenetic transmission inference.

Competing interests

The author declares that they have no competing interests.

Use of Artificial Intelligence (AI)

The author declares that they did not use generative AI to come up with new ideas for their review.