Summary
Background
Malaria is a mosquito borne disease caused by parasites of the genus Plasmodium. In 2023, the United States (US) experienced nine cases of autochthonous Plasmodium vivax malaria transmission; seven in Florida, one in Texas, and another in Arkansas. These were the first autochthonous cases since 2003 when a cluster was identified in Florida. The aim of this study was to genetically characterize the implicated P. vivax isolates in order to complement epidemiologic investigations of these cases.
Methods
A custom Illumina AmpliSeq sequencing panel capturing 495 amplicons was designed. This panel was used to ascertain whether these 2023 cases were related, and assess if they were associated with a single or separate introduction events. Sequence data were hierarchically clustered and a Naïve Bayes classification approach was used to assign genotypes to a probable geographic origin based on 113 ‘geo-informative’ SNPs captured by the panel. Genotypes associated with the 2023 Arkansas, Texas, and Florida cases were clustered alongside those sequenced from archived blood samples from the 2003 Florida case-patients, a set of reference strains, and other travel-associated specimens. Microsatellite analysis was performed on a subset of samples from these autochthonous cases to complement the AmpliSeq analysis.
Findings
The 2023 autochthonous Florida cases were genetically linked as were the 2003 Florida cases. The 2023 and 2003 Florida clusters were genetically distinct, and the two Florida clusters were distinct from the 2023 Texas and Arkansas cases, which were also distinct from each other. These genotypes classified to the Central or South American region using the Naïve Bayes classifier, including those from the 2003 cluster.
Interpretation
These data support that at least three distinct P. vivax introduction events in the US in 2023, involving parasites possessing genetic signatures consistent with Central or South America.
Funding
This work was supported by the National Center for Emerging and Zoonotic Infectious Diseases at the US Centers for Disease Control and Prevention.
Keywords: Malaria, Autochthonous, Plasmodium vivax, USA, Transmission, Sequencing, Genotyping, AmpliSeq, Barcoding
Research in context.
Evidence before this study
Autochthonous Plasmodium vivax malaria transmission has been reported sporadically in the United States (US) since being considered malaria-free in 1951. In 2023, nine cases of locally acquired P. vivax malaria were reported to the US Centers for Disease Control and Prevention (CDC); 7 from Florida, one from Texas, and one from Arkansas, representing the first autochthonous US cases since a cluster reported in Florida in 2003. Public health investigations of the 2023 cases ensued. Genetic characterization of the parasites involved would complement investigations by addressing key questions on the parasites' origin and their genetic relationship to each other. PubMed article searches using the terms “Plasmodium vivax barcoding” (34 results) and “Plasmodium vivax next generation sequencing” (56 results) with no date or language restrictions, revealed that P. vivax molecular barcoding approaches varied by their evaluation settings, with efficacy often reported in the context of a particular study site or region. In this study, the parasites’ origin was suspected though not confirmed genetically, and the possibility that future P. vivax introductions might originate from anywhere across its endemic range cannot be excluded. Thus, P. vivax genotyping methods for use in this context would ideally remain efficacious when applied to strains of diverse origins. The need to identify close genetic kinship relationships was also a key consideration when selecting genetic markers. A set of markers was compiled, influenced by P. vivax genetic barcoding panels from the scientific literature possessing the desired characteristics. These markers informed the design of a custom Illumina AmpliSeq sequencing panel.
Added value of this study
This study describes evaluation and deployment of this AmpliSeq panel to complement investigations of the autochthonous P. vivax transmission events that occurred in the US in 2023. The panel was also applied to P. vivax-positive blood samples from a cohort of US patients with diverse travel histories, including travel to various parts of Asia, Africa, Oceania, and Latin America. The panels’ efficacy in assigning P. vivax strains to a geographic origin was confirmed; congruency between travel destinations and origins assigned based on genetic signatures was high. A close genetic relationship among parasites causing the 2023 Florida infections was confirmed, and the 2023 Florida parasites were distinct from those associated with the Texas and Arkansas cases. The Texas and Arkansas parasites were also distinct from each other. Archived blood samples from seven patients implicated in the 2003 outbreak were analyzed, demonstrating for the first time that these parasites shared close genetic kinship, and were therefore likely associated with a single introduction. Ultimately, the panel proved efficacious at addressing crucial questions that arose during investigations of these high-profile P. vivax transmission events.
Implications of all the available evidence
At least three distinct P. vivax introductions occurred in the US in 2023 and the parasites involved in the 2003 and 2023 autochthonous P. vivax transmission events possessed genetic signatures consistent with a Central or South American origin. Collectively, the evidence supports that certain Southeastern US jurisdictions have been receptive to P. vivax malaria introductions from certain regions for many years, and likely remain vulnerable to local P. vivax transmission.
Introduction
Malaria is a mosquito borne disease caused by parasites of the genus Plasmodium. Several Plasmodium species cause human malaria including Plasmodium falciparum, P. vivax, Plasmodium ovale curtisi, P. ovale wallikeri, Plasmodium knowlesi, and Plasmodium malariae. Each is a significant cause of morbidity and mortality across their endemic range, though P. falciparum constitutes the greatest threat to human health.1 While less deadly than P. falciparum, P. vivax causes a debilitating and potentially life-threatening form of malaria, with the potential for illness relapse as a consequence of dormant liver infection.2 P. vivax is the most widely distributed cause of human malaria; its endemic range spans most tropical and subtropical parts of the world (e.g., the Americas, Mediterranean, Southeast Asia, and the Asia–Pacific region) with the exception of much of Sub-Saharan Africa,3,4 due to the high prevalence of the Duffy negative blood phenotype among peoples native to this region.5 Approximately 7 million P. vivax malaria cases were reported worldwide in 2022, comprising around 2.8% of all 249 million malaria cases.4
Currently, most Central and South American P. vivax malaria cases occur in Venezuela, Brazil, and Colombia, though cases occur elsewhere within the region, including Nicaragua, Honduras, Costa Rica, and Panama.4 Endemic P. vivax transmission was reported to the World Health Organization (WHO) from as far north on the American continent as Mexico.4 The United States (US) has been considered malaria-free since 1951, following the successful National Malaria Eradication Program (commencing 1947) involving vector control via widespread dichlorodiphenyltrichloroethane (DDT) application to households in parts of the Southeast.6 Despite these efforts, malaria vectors remain endemic in the US (e.g., Anopheles freeborni, An. quadrimaculatus, An. crucians) and mosquito-borne malaria transmission has led to more than 60 small-scale US transmission events over the past 50 years.7 Prior to 2023, autochthonous US cases of mosquito-borne P. vivax malaria were confirmed in 2003, when an epidemiologically defined cluster of eight P. vivax malaria cases was identified in Palm Beach County, Florida.8
On May 18th 2023, Florida Department of Health (FDOH) requested assistance from the US Centers for Disease Control and Prevention (CDC) to confirm malaria in a patient with no history of travel to malaria-endemic areas9 and a diagnosis of P. vivax malaria was confirmed. Six additional autochthonous P. vivax malaria cases were reported to CDC by FDOH between June 19 and July 17, totalling seven patients residing within a 4-mile radius. Independently, on June 7, a hospital in Texas requested CDC's assistance to verify a malaria case in a patient lacking a history of travel or other risk factors, and another P. vivax infection was confirmed.9 A ninth local P. vivax malaria case was diagnosed in September 2023 in Arkansas. The temporal and spatial proximity of the seven Florida cases suggested they were linked to the same introduction event, while the Texas and Arkansas cases were considered unrelated.
The Division of Parasitic Diseases and Malaria at CDC sought to develop a P. vivax genotyping tool based on next generation sequencing technologies, to complement epidemiologic investigations of autochthonous US malaria transmission and explore whether the 2023 malaria cases were genetically closely related.
Methods
Samples and controls
State Public Health Laboratories (SPHLs) in the United States receive post-diagnostic whole-blood samples from patients diagnosed with malaria by local healthcare providers within their jurisdictions. Post-diagnostic whole blood samples included in this study were deidentified and voluntarily transferred to CDC by SPHLs for downstream analysis. Where available, collection dates for blood samples are provided in File S1, Tab A. DNA was extracted from EDTA-treated whole blood using a QIAamp DNA Mini Kit (Qiagen, Germany) following the manufacturer's instructions, using 200 μL of blood as input and an elution volume of 200 μL. Extracts were stored at −20 °C prior to being subjected to AmpliSeq sequencing. A subset of samples were subjected to microsatellite analysis. In addition to specimens from the autochthonous malaria cases from Florida, Texas, and Arkansas (referred to herein as domestic specimens), analysis included P. vivax positive blood samples from travel-associated malaria cases collected from 2012 to 2023, submitted from partnering US State Public Health laboratories (File S1, Tab A). Sequencing was performed for 203 individual samples; 194 historic or travel related P. vivax-positive bloods (including 7 from the 2003 Florida cases), plus 9 samples from the 2023 Arkansas, Texas and Florida cases (one from Texas, one from Arkansas, seven from Florida). To evaluate sequencing reproducibility, repeat sequencing was performed for a total of 15 reproducibility replicates (File S1, Tab A). Samples were sequenced across six AmpliSeq libraries, each accompanied by at least one positive control (DNA extracted from blood containing P. vivax strain Pv-Br7 or Salvador I), one negative template control (PCR grade water), and one negative extraction control comprised of either parasite negative human blood or cultured P. falciparum strain Dd2 or D6. Six sets of controls were generated, and a subset is shown in File S1, Tab A to demonstrate the types of controls sequenced (n = 10), including data generated for different stored aliquots of the same control (e.g., P. vivax Salvador I) or one result to represent repeats of certain negative controls (e.g., repeats of P. falciparum D6 and Dd2). Including the 15 reproducibility replicates and 10 controls, 228 sets of paired-end fastq files are included in File S1, Tab A.
Ethics approval
Ethics approval for the use of clinical specimens was obtained from CDC (project determination accession number CGH-LSDB-3/22/23-fe44d).
Microsatellite analysis
P. vivax associated with the 2023 Florida and Texas autochthonous cases, and parasites implicated in the 2003 Palm Beach outbreak, were genotyped by fragment-length analysis of eight microsatellites. DNA was extracted from EDTA-treated whole blood as described above. DNA extracts were subjected to eight PCRs, each targeting a different microsatellite locus10, 11, 12: MS6, MS038, Pv3.502, Pv10.29, Pv11.162, Pv12.335, Pv14.185, and Pv14.297. Methods for preparing these PCRs are provided in File S2 (Microsatellite PCR methods). Amplicon fragments were separated by capillary electrophoresis on an ABI 3500xL Genetic Analyzer (Thermo-Fisher), according to the manufacturer's instructions. Fragment sizes were determined by analysing electropherogram files using GeneMarker® software (Softgenetics).
AmpliSeq panel design
The Ampliseq panel captures 495 amplicons sequenced via two oligomer pools, with some overlap among amplicons. Amplicons are distributed across all 14 P. vivax chromosomes (Fig. 1) and include markers from Kattenberg et al.,14 plus those from Siegel et al.,15 Baniecki et al.13 (i.e., the Broad Institute panel), and a selection of additional markers (i.e., part of the genes encoding the apical membrane protein, Duffy binding protein, merozoite surface protein, and circumsporozoite protein). The panel captures all unique markers from the Geo30, Geo50, and Geo55 panels (totalling 113 SNPs) for characterizing the P. vivax parasites by geographic region.15,16 Oligos targeting these markers were identified relative to the PvP01 reference genome (BioSample: SAMEA2142821), and the design was finalized with guidance from the Illumina Concierge team (Illumina, San Diego, USA).
Fig. 1.
Approximate position of AmpliSeq panel markers relative to each of 14 P. vivax chromosomes. Approximate position of the Broad institute markers13 (red circle), putative drug resistance markers from Kattenberg et al.14 (olive triangle), the Geo30, Geo50, and Geo55 markers15,16 (green square), microhaplotypes described by Siegel et al.15 (pink square with “X”), and other custom targets (blue “+” sign). Precise marker locations relative to the PvP01 reference genome assembly (Version 1, BioSample: SAMEA2142821) are provided in File S3, Tab ‘design.bed’.
AmpliSeq sequencing
DNA extracts were quantified using a Qubit High sensitivity DNA kit (Invitrogen, USA), and 50–150 ng of DNA was used as input for the AmpliSeq amplification reaction. Targets were amplified in two oligonucleotide pools (File S3, Tab: ‘design.bed’) using the following thermocycling conditions: 99 °C for 2 min, followed by 21 cycles of 99 °C for 15 s and 60 °C for 8 min. Amplicons from both pools were combined in a 1:1 ratio prior to library preparation using an AmpliSeq Library PLUS for Illumina kit and AmpliSeq CD Indexes (Illumina), following the manufacturer's guidelines. Library concentration was assessed using the Qubit High sensitivity DNA kit and fragment size was determined on a TapeStation using High Sensitivity D5000 reagents. Quantified libraries were diluted to 2 nM with low Tris–EDTA buffer, and pooled. The diluted library pool was prepared to contain a 1% PhiX spike-in (Illumina) and denatured using NaOH. The library was sequenced on the Illumina MiSeq platform using a Miseq Reagent Kit v2 (Illumina, USA) resulting in 2 by 250 base pair paired end reads.
Clustering
Genotypes were constructed from paired-end fastq files using Module 1 of the previously described CYCLONE workflow.17,18 This workflow was developed for clustering Cyclospora amplicon data17; modifications were made to the reference files as described in the CYCLONE manual (see Barratt et al.17) and File S2 (Analysis of AmpliSeq data). The final output of CYCLONE Module 1 is a haplotype data sheet (HDS; a tabular format for displaying genotypes17,19). Using custom code, genotypes were extracted from 76 reference datasets from P. vivax strains of known geographic origin (from GenBank and MalariaGen20) and added to the HDS (File S1, Tab B) for comparison. A genetic distance matrix was computed from the resultant HDS using Barratt's heuristic.19 If a partial genotype is obtained from a sample (e.g., due to poor DNA quality or low parasite load), Barratt's heuristic attempts to impute distance values for missing markers to compensate. However, this requires investigators to provide the algorithm with a set of minimum data requirements to ensure that imputation is attempted only for sufficiently complete genotypes.19 CYCLONE Module 1 generates text files containing the genotype of each sample, represented as a series of BLASTN results. A complete P. vivax genotype generated using the present AmpliSeq panel has a text file size ranging from 238 to 250 kilobytes. For this study, only genotype text files of 200 kilobytes or larger (i.e., >83% complete) were added to the HDS for genetic distance computation.
A genetic distance matrix was computed from the HDS using a python implementation of Barratt's heuristic (https://github.com/Joel-Barratt/Eukaryotpying-Python19), with modifications to improve clustering performance and speed (File S2–Genetic distance computation). This matrix was hierarchically clustered using an R implementation of Ward's method21 (AGNES R package). The resulting hierarchical tree was rendered and annotated using ggtree.22 To define discrete clusters of closely related P. vivax within the tree, a genetic distance (delta) threshold for linking genotypes was computed using a previously described method,23 with some modifications (File S2–Modified tree dissection method). PARNAS24 was used to dissect the tree with a radius setting equal to the delta threshold as described elsewhere.23
Geographic classification
A standalone implementation of a 'Bi-Allele Likelihood’ (BALK) Naïve Bayes classifier16 was trained using 1012 P. vivax genomes of diverse geographic origins from MalariaGen.20 Specifically, a BALK classification model was trained using 113 SNPs across the union (GeoCombined) of the Geo30, Geo50, and Geo55 panels.15,16 Illumina reads were quality filtered and trimmed using BBDUK (minimum length = 50, minimum PHRED quality score = 30). Reads were aligned to the GRCh38 human reference genome (GenBank: GCA_000001405.29) using bowtie2. Samtools was used to extract reads not mapping to this reference. Using bowtie225 remaining reads were aligned to AmpliSeq reference sequences capturing the 113 target SNPs (Files S3, Tab ‘design.bed’ and File S4). Aligned reads were merged and sorted using samtools.26 GATK427 CleanSam was used to soft-clip alignments extending beyond the reference and generate GVCFs for each sample via HaplotypeCaller mode. GVCFs were combined and genotyped with GATK4 CombineGVCFs and GenotypeGVCFs (with the –include-non-variant-sites flag), respectively. A sample was excluded from classification if it failed to obtain coverage for at least 57 of the 113 SNP sites (∼50% of sites): an empirically-selected cutoff based on user experience. For comparison, the classifiers performance was also assessed when trained on the Geo33, Geo50, and Geo55 SNP panels individually (File S2–Training datasets for classification).
Travel data
Travel information for US malaria cases was requested from US States submitting malaria-positive samples to CDC for molecular surveillance purposes, and was captured on standardized sample submission forms accompanying physical samples. Notably, complete travel histories were not always provided, or were incomplete and possibly inaccurate. When available, travel histories were compared to their respective geographic classifications to assess their concordance. For cases where a travel history was not provided, these data were considered ‘missing’. No attempt was made to follow up on missing travel data within the scope of this study, except where an autochthonous case was suspected.
Institutional review board statement
The use of clinical samples and epidemiologic data in this study was reviewed by CDC and was conducted consistent with applicable federal law and CDC policy, and in accordance with human research protection procedures (project determination CGH-LSDB-3/22/23-fe44d).
Role of the funding source
This work was funded by the National Center for Emerging and Zoonotic Infectious Diseases at the US Centers for Disease Control and Prevention in support of malaria preparedness. Only the authors played a role in the design and writing of this study, and in the decision to submit it for publication.
Results
Microsatellite analysis
Five of 7 samples from the 2003 Florida P. vivax outbreak produced an identical microsatellite profile. Two samples (FL_2003_01 and FL_2003_07) differed only at marker 14.297 by 3 bases (Table 1). Similarly, profiles obtained for the seven 2023 Florida specimens were largely similar with minor differences observed in 3 specimens. One of these differed at marker 14.297 and 2 differed at marker MS038. The Florida 2003 and Florida 2023 profiles were dissimilar. The microsatellite profile obtained from the 2023 domestic Texas case was distinct from the Florida autochthonous cases with their consensus differing at 7 of 8 markers. A microsatellite profile generated for a 2023 Texas imported case, diagnosed in close temporal and geographic proximity to the domestic Texas case, was also distinct from the Texas 2023 domestic case and all Florida cases, differing at 4 or more markers (Table 1).
Table 1.
Microsatellite amplicon lengths as determined by capillary electrophoresis.
| Repeat motif |
TCC (chr 11) |
CATA (chr 6) |
AACGGATG (chr 3) |
AT (chr 10) |
ATAC (chr 11) |
AT (chr 12) |
AT (chr 14) |
AAG (chr 14) |
|
|---|---|---|---|---|---|---|---|---|---|
| Marker | MS6 allele #1 | MS038 allele #1 | MS038 allele #2 | 3.502 allele #1 | 10.29 allele #1 | 11.162 allele #1 | 12.335 allele #1 | 14.185 allele #1 | 14.297 allele #1 |
| Amplicon lengths from samples related to the Florida 2003 cases | |||||||||
| FL_2003_01 | 234 | 330 | 334 | 159 | 114 | 184 | 170 | 270 | 198 |
| FL_2003_02 | 234 | 330 | 334 | 159 | 114 | 184 | 170 | 270 | 201 |
| FL_2003_03 | 234 | 330 | 334 | 159 | 114 | 184 | 170 | 270 | 201 |
| FL_2003_04 | 234 | 330 | 334 | 159 | 114 | 184 | 170 | 270 | 201 |
| FL_2003_05 | 234 | 330 | 334 | 159 | 114 | 184 | 170 | 270 | 201 |
| FL_2003_06 | 234 | 330 | 334 | 159 | 114 | 184 | 170 | 270 | 201 |
| FL_2003_07 | 234 | 330 | 334 | 159 | 114 | 184 | 170 | 270 | 198 |
| Amplicon lengths from samples related to the Florida 2023 cases | |||||||||
| FL_2023_01 | 252 | 200 | NA | 141 | 118 | 180 | 178 | 268 | 201 |
| FL_2023_02 | 252 | 200 | NA | 141 | 118 | 180 | 178 | 268 | 201 |
| FL_2023_03 | 252 | 200 | NA | 141 | 118 | 180 | 178 | 268 | 198 |
| FL_2023_04 | 252 | 198 | NA | 141 | 118 | 180 | 178 | 268 | 201 |
| FL_2023_05 | 252 | 198 | NA | 141 | 118 | 180 | 178 | 268 | 201 |
| FL_2023_06 | 252 | 200 | NA | 141 | 118 | 180 | 178 | 268 | 201 |
| FL_2023_07 | 252 | 200 | NA | 141 | 118 | 180 | 178 | 268 | 201 |
| Amplicon lengths from samples related to the Texas 2023 domestic case | |||||||||
| TX_2023_01a | 210 | 274 | 278 | 159 | 114 | 196 | 164 | 268 | 198 |
| TX_2023_02a | 210 | 274 | 278 | 159 | 114 | 196 | 164 | 268 | 198 |
| Amplicon lengths from samples related to the Texas 2023 imported case | |||||||||
| TX_2023_03b | 249 | 274 | NA | 159 | 118 | 196 | 164 | 272 | 192 |
| TX_2023_04b | 249 | 274 | NA | 159 | 118 | 196 | 164 | 272 | 192 |
Notes: Two alleles were detected at marker MS6 for some specimens. For isolates where one allele was observed the column MS038 allele #2 will indicate ‘NA’. Fragment lengths differing from the consensus of each group are highlighted in bold.
Duplicate genotypes from the same patient (technical replicates); the 2023 domestic case from Texas.
Duplicate genotypes from the same patient (technical replicates); the 2023 imported case from Texas.
AmpliSeq genotyping
Hierarchical clustering
All negative control files yielded insufficient data for clustering. Fastq files from 19 clinical samples also contained insufficient data for clustering. A distance matrix (File S1, Tab C) was computed for the remaining 204 samples (including 15 reproducibility replicates and 5 positive P. vivax controls: 1 Pv-Br7 and 4 repeats of Salvador I), plus genotypes extracted from 76 reference genomes (n = 280 genotypes). A delta threshold of 0.2642226 was computed. Corresponding reproducibility replicates and repeat sequencings of Salvador I were separated by a genetic distance below 0.2642226 (indicating genetic linkage), as was the genotype extracted from the Salvador I reference genome (GenBank: GCA_000002415.2, File S1, Tab C). A hierarchical tree (Fig. 2) was generated from a second matrix (File S1, Tab D) that excluded the 15 reproducibility replicates (leaving 189 genotypes sequenced here and those from the 76 reference genomes, n = 265). Using a radius setting of 0.2642226, PARNAS dissected this tree into 183 clusters (File S1, Tab E).
Fig. 2.
Hierarchical tree showing clustered P.vivax genotypes. This tree contains 265 genotypes—189 sequenced for this study and 76 extracted from published reference data (File S1, Tabs C and D). Branches are colour coded according to the geographic classifications assigned to genotypes on the corresponding branch tips. Eight classification categories are highlighted, and are largely congruent with the hierarchical tree structure; Africa (AF)—red branches, East Asia (EAS)—blue branches, East Southeast Asia (ESEA)—green branches, Latin America (LAM)—purple branches, Malaysian archipelago/Maritime Southeast Asia (MSEA)—orange branches, Oceania (OCE)—gold branches, West Asia (WAS)—pink branches, and West Southeast Asia (WSEA)—grey branches. Genotypes assigned to the MSEA category were obtained solely from published reference data—no genotypes sequenced here obtained a MSEA classification. The map (bottom left) shows these regions coloured according to the branches. Coloured bars encircling this tree reflect distinct clusters, where adjacent bars of the same colour reflect specimens assigned to the same cluster by PARNAS using a radius of 0.2642226. Peripheral bar colours have no additional meaning. A total of 183 clusters are shown including 160 singletons (clusters comprised of one genotype). Genotypes from notable cases are highlighted, including a Texas (TX) traveller with a pre- and post-treatment genotype, an Ethiopia traveller, the TX imported case where the patient subsequently travelled to Minnesota (MN), the TX 2023 domestic case, the 2023 Arkansas (AR) domestic case, the Florida (FL) 2023 domestic cases, and the FL 2003 domestic cases. A cluster comprising repeat sequencings of Salvador I strain is also shown noting that one genotype in this cluster was extracted from the Salvador I reference genome on NCBI. Clusters marked with two asterisks include those comprising samples submitted from two or more US jurisdictions (States). Those marked with one asterisk include genetic clusters of samples submitted from a single US State. The cluster containing the sample from the imported TX case (two asterisks) includes samples submitted from jurisdictions in addition to TX and MN.
Samples from the 2023 autochthonous Florida cases were genetically linked and clustered closely (delta <0.2642226) as were samples from the 2003 Florida domestic cases (Fig. 2). The 2023 and 2003 Florida clusters were genetically distinct (Fig. 2). These two Florida clusters were also distinct from both 2023 Texas cases (domestic and imported), which were also distinct from one another. The genotype obtained from the autochthonous Arkansas case was distinct from the 2023 Texas imported and domestic cases, and the Florida 2003 and 2023 domestic cases. Other notable clustering results include those obtained for a Texas travel-related case where pre- and post-treatment genotypes clustered closely (delta <0.2642226). Additionally, a patient with imported malaria initially sought care in Texas but received treatment that did not clear the parasite. The patient later travelled to Minnesota and was diagnosed with malaria again and was treated. Genotypes obtained from both blood samples collected on separate occasions (in different States) clustered together. Similarly, two genotypes associated with travel to Ethiopia clustered together and were generated from two samples collected from the same patient. Additionally, eight clusters of tightly-linked genotypes of a likely Latin American origin were detected among specimens from patients with travel to malaria endemic areas (Fig. 2—asterisked clusters).
Geographic classification
When trained on the GeoCombined SNP panel, the BALK classifier displayed superior performance compared to the Geo30, Geo50, and Geo55 panels (File S2–Figure S2). Thus, subsequent classifications utilized a GeoCombined model. Of the 218 sequenced samples (excluding positive and negative controls), four obtained insufficient coverage at more than 50% of the 113 SNPs. This left 214 samples for classification, of which 97 had available travel data and were not from the nine autochthonous cases (Table 2). Of these 97 samples, 95 classified to a geographic region congruent with their respective laboratory-reported travel histories (98% concordance), noting that three of these 95 were replicates from the same patient (Table 2). These 97 samples covered diverse geographic regions and classified to seven of the eight geographic categories (Table 2), excluding Maritime Southeast Asia (MSEA) (Fig. 2, Table 2, File S2—Table S2). Of the two discordant results, one specimen had a reported travel history to Brazil but obtained an East-Southeast Asian classification (ESEA), and the second had a reported travel history to the US Virgin Islands and obtained a West Asian (WAS) classification. Specimens from domestically acquired P. vivax cases (i.e., Florida 2023, Florida 2003, Texas 2023, and the Arkansas 2023 case) as well as the Texas 2023 imported case noted above, each obtained a LAM classification. Overall, classifications were consistent with clustering results (Fig. 2).
Table 2.
Travel-related P. vivax malaria cases with a reported travel history that was congruent with their geographic classification.b
| Reported/preliminary travel historiesa | Classification | Sample countsc | Total counts |
|---|---|---|---|
| Africa | AF | 1 | Africa (AF) Total: 21 |
| Burundi | AF | 1 | |
| Congo | AF | 1 | |
| Ethiopia | AF | 10 (9 cases) | |
| Ethiopia and Dubai | AF | 1 | |
| Ethiopia and Eritrea | AF | 1 | |
| Kenya | AF | 1 | |
| Mali | AF | 1 | |
| Sudan | AF | 2 | |
| Uganda and Kenya | AF | 1 | |
| Uganda, Kenya, Zambia, and South Africa | AF | 1 | |
| Amazon | LAM | 1 | Central or South America (LAM) Total: 35 |
| Brazil | LAM | 3 | |
| Central America | LAM | 1 | |
| Colombia | LAM | 1 | |
| Cuba, Central America, and Mexico | LAM | 2 (1 case) | |
| Ecuador | LAM | 4 | |
| Guatemala | LAM | 4 | |
| Honduras | LAM | 3 | |
| Mexico | LAM | 4 | |
| Mexico and Cuba | LAM | 2 | |
| Nicaragua | LAM | 1 | |
| Nicaragua and Mexico | LAM | 1 | |
| Peru | LAM | 3 | |
| Peru and Colombia | LAM | 1 | |
| Peru, Columbia, and Panama | LAM | 1 | |
| Venezuela | LAM | 3 | |
| Fiji | OCE | 1 | Oceania (OCE) Total: 8 |
| Indonesia | OCE | 1 | |
| Papua New Guinea | OCE | 4 | |
| Solomon Islands | OCE | 2 | |
| Afghanistan | WAS | 3 | West Asia (WAS) Total: 24 |
| India | WAS | 12 | |
| India and Pakistan | WAS | 1 | |
| Pakistan | WAS | 5 | |
| Pakistan and Afghanistan | WAS | 3 | |
| South Korea | EAS | 2 (1 case) | East Asia (EAS) Total: 2 |
| Cambodia | ESEA | 2 | East Southeast Asia (ESEA) Total: 2 |
| Myanmar | WSEA | 1 | West Southeast Asia (WSEA) Total: 2 |
| Thailand | WSEA | 1 | |
| Total samples where the travel history matched the classification category: | 94 samples (91 cases) | ||
Notes: No samples of the MSEA category were identified among samples analyzed here.
Preliminary travel histories reported on sample submission forms without further verification. Some travel histories reflected travel to multiple countries, and each unique set of countries reported across all cases has a discrete row even if some rows contain one or more countries that intersect.
The two classifications that were incongruent with the preliminary travel histories are not listed in this table.
Number of samples with this preliminary travel history. Note that three cases had two classification results each as these were among our 15 reproducibility replicates.
Discussion
In 2023, the US identified nine cases of autochthonous P. vivax transmission across three states (FL, TX and AR), confirming that certain US jurisdictions are receptive to P. vivax malaria importation and vulnerable to local transmission. The AmpliSeq panel and accompanying analytical approaches described here, suggest that three separate P. vivax introductions (FL, TX, and AR) in 2023 resulted in autochthonous US malaria transmission, and the parasites have genetic signatures consistent with a Latin America origin. Furthermore, these methods facilitated comparisons between present cases and notable historic cases (e.g., the outbreak of autochthonous malaria transmission in Florida, 2003). Ultimately, establishment of next-generation sequencing and bioinformatic capacity within the Division of Parasitic Diseases and Malaria at CDC facilitated the rapid implementation of methods and analysis pipelines in support of these 2023 P. vivax case investigations. While not a focus of the present study, the putative genetic determinants of drug resistance from Kattenberg et al.14 captured by the present AmpliSeq panel may be applicable to investigations of P. vivax drug resistance should the need arise in the future.
Autochthonous US malaria transmission represents a high-profile situation, so the techniques described here were evaluated via multiple routes. For example, epidemiologically linked cases were genetically linked by microsatellite analysis, and then by the AmpliSeq-based clustering approach; thus, three independent analyses supported similar conclusions. The epidemiologic utility of the clustering analysis was highlighted further by results comparing the 2023 Texas domestic to the Texas imported case, which shared a temporo-spatial relationship, raising suspicion that they were related. However, clustering supported that these cases were unrelated—a conclusion supported by the microsatellite analysis. It also bears noting that Muneer et al.28 sequenced whole genomes from blood taken from four of the seven 2023 Florida case-patients and confirmed genetic linkage of the associated parasites in addition to their LAM genetic signature.28 Repeat sequencing of reproducibility replicates showed consistent clustering among corresponding replicates supporting reproducible amplification across sequencing runs. Finally, controlling for investigator bias is a key but often neglected consideration when developing molecular tools to support epidemiologic investigations: investigator expectations may bias interpretation of genetic clustering outcomes.29 Automating certain decision-making processes can alleviate bias. Here, the automatically computed delta threshold served as the PARNAS radius setting for tree dissection, a process that might have otherwise been manually performed by investigators with knowledge of the associated epidemiologic links.
Limitations
The BALK classifier16 trained on the GeoCombined panel was efficacious, where 98% congruence was observed between travel histories and classifications. Two of 97 classifications were discordant which could prompt further epidemiological investigation. One sample was associated with travel to US Virgin Islands but was classified to WAS. As malaria is not endemic to the US Virgin Islands, the infection likely originated elsewhere, presumably West Asia. For the discordant Brazil travel case that classified to ESEA, 96% of SNPs required for classification were sequenced, yet roughly half of those positions were heterozygous. The sample also lacked sufficient data for clustering. The partial yet complex nature of this genotype likely impacted classification accuracy. In this study, travel histories were extracted directly from sample submission forms, though as a nationally notifiable disease, malaria epidemiological data are also submitted to CDC via the National Malaria Surveillance System (NMSS) and the National Notifiable Diseases Surveillance System (NNDSS). The NMSS and NNDSS epidemiological data are subjected to independent verification via thorough public health investigations. This verification process was not completed within the scope of the present study, so the travel data presented here should be viewed as unverified and preliminary. This may account for the discordant geo-classification results.
The 2023 and 2003 Florida clusters had genetic signatures consistent with LAM parasites though they clustered separately, as expected based on their temporal separation. The results obtained for the 2003 Florida cases (sequenced here for the first time) show that P. vivax introduction from LAM occurred previously and again in 2023. While not a focus here, an attempt was made to sequence P. vivax from three PCR-positive A. crucians mosquitoes trapped in Florida in 2023 (File S2—Supplementary Results). Unfortunately, these specimens yielded insufficient data for clustering, though one genotype obtained a LAM classification. This further highlights the risk for future introductions from LAM and provides direct evidence that A. crucians is a permissible P. vivax vector in the US. In relation to the geo-classification analysis, a limitation of this study is the lack of resolution presented within the LAM category. The main obstacle to increased geographic resolution in LAM is the limited availability of P. vivax genomic data with only a few countries represented at the time this work was performed (namely Mexico, Brazil, Colombia, and Peru). However, recent work does provide evidence for geographic sub-structuring of P. vivax populations within the LAM region.30 As such, the availability of P. vivax genome data from other LAM countries may provide opportunities to improve the resolution of this geo-classification method in within LAM.
As with any hierarchical clustering approach, outcomes depend heavily on the test populations composition; tree structures can change slightly with the addition or subtraction of specific genotypes. The efficacy of the BALK classifier may also be limited by available reference data, as certain P. vivax-endemic countries are more widely sampled than others (File S2—Table S2). Global P. vivax populations are likely in flux as human activities (e.g., travel and migration) alter the P. vivax population structure in some regions. To support clustering and classification accuracy, a database of geographically diverse historic and novel genotypes via routine sequencing of field and travel-associated strains, will be needed. This could facilitate detection of changes in population structure, allowing laboratory investigators to continually evaluate workflow efficacy, and adjust workflow parameters accordingly.
Conclusions
Next-generation sequencing is a powerful complement to outbreak investigations, where the additional genetic insight could improve the efficacy of public health interventions aiming to limit subsequent malaria introductions or spreading (e.g., by identifying at risk areas in need of resource allocation). Here, a detailed genetic analysis of P. vivax associated with autochthonous US malaria cases is presented, facilitated by AmpliSeq technology. Our analysis further supports that P. vivax malaria was introduced to the US at least three separate times in 2023, and that the parasites involved were likely of Latin American (LAM) origin. Parasites associated with the 2003 Florida cluster also classified to LAM. Additionally, eight tightly associated genetically-defined clusters with a LAM genetic signature were defined from a series of imported malaria cases. Taken in conjunction with increased imported malaria diagnoses at Southern US jurisdictions31 these events highlight the potential for future P. vivax introductions from LAM compared to other P. vivax endemic areas. Ultimately, there is a need for sustained US malaria preparedness, including continued malaria sequencing capacity at State and Federal public health agencies to ensure robust investigations and responses in the event of future autochthonous cases of US malaria transmission.
Contributors
Joel Barratt: inception, design, direction, methods, data analysis, code, interpretation, coordination, preparation of original draft, verified the data, had access to raw data, and had final responsibility for the decision to submit for publication.
David Jacobson: design, methods, data analysis, code, interpretation, review of drafts.
Marko Bajic: design, methods, laboratory work, interpretation, review of drafts.
Edwin Pierre Louis: design, methods, laboratory work, interpretation, review of drafts.
Julia Kelley: methods, laboratory work, interpretation, review of drafts.
Dhruviben S. Patel: design, methods, code, interpretation, review of drafts.
Ira Goldman: methods, lab work, review of drafts.
Kimberly Mace: epidemiologic methods, data analysis, review of drafts.
Zhiyong Zhou: methods, design, and data analysis, review of drafts.
Ya Ping Shi: design, direction, and interpretation, review of drafts.
Alison Ridpath: epidemiologic methods, data analysis, review of drafts.
Christina Carlson: sample processing, review of drafts.
Alice Sutcliffe: sample processing, review of drafts.
Qiana Butler: sample processing, review of drafts.
Andrea Morrison: epidemiologic investigations, review of drafts (Florida).
Danielle Stanek: epidemiologic investigations, review of drafts (Florida).
Kelly Tomson: epidemiologic investigations, review of drafts (Florida).
Andrew Cannons: laboratory investigations, review of drafts (Florida).
Carina Blackmore: epidemiologic investigations, review of drafts (Florida).
Susan Rollo: epidemiologic investigations, review of drafts (Texas).
Chun Wang: laboratory investigations, review of drafts (Texas).
Rashmi Tuladhar: laboratory investigations, review of drafts (Texas).
Brooke Clemons: sample processing, review of drafts (New York Wadsworth).
Susan Madison-Antenucci: sample processing, review of drafts (New York Wadsworth).
Kimberly Mergen: sample processing, review of drafts (New York Wadsworth).
Jennifer White: epidemiologic investigations, review of drafts (New York State).
Mike Antwi: epidemiologic investigations, review of drafts (New York City).
Laura Rothfeldt: epidemiologic investigations, review of drafts (Arkansas).
Jennifer N. Shray: laboratory investigations, review of drafts (AR).
Ashleah Courtney: malaria diagnosis, review of drafts (AR).
Bobby Boyanton: malaria diagnosis, review of drafts (AR).
Katelyn Lazenby: laboratory investigations, review of drafts (AR).
Stephen Hedges: epidemiologic investigations, review of drafts (AR).
Molly Freeman: administration, direction, review of drafts.
Yvonne Qvarnstrom: sample preparation, specimen handling, analysis, drafts.
Brian Raphael: inception, coordination, administration, direction, review of drafts.
Data sharing statement
Sequence data have been deposited in the NCBI Sequence Read Archive (SRA) under BioProject PRJNA1092573. All code used to perform the BALK classifier evaluation variant calling pipelines, modified CYCLONE workflow, and all custom scripts have been made available via GitHub: https://github.com/CDCgov/pvivax_ampliseq_analysis).
Editor note
The Lancet Group takes a neutral position with respect to territorial claims in published maps and institutional affiliations.
Declaration of interests
We declare no competing interests.
Acknowledgements
We acknowledge the support of the National Center for Emerging and Zoonotic Infectious Diseases’ Division of Parasitic Diseases and Malaria (U.S. Centers for Disease Control and Prevention) for their support of this work. Special thanks to the various U.S. State health departments that submitted any specimens to CDC that were included in this analysis (this includes CA, KS, NY, FL, NV, MA, AZ, AR, TX, MN, NJ, OR, PA, MO, KY, WI, SC, MS, WA, RI, GA, VA, MI, MT, WV, NC, AL, DC, UT, IL, LA). We acknowledge the New York City Department of Health and Mental Hygiene Bureau of Communicable Disease's Disease Investigation Unit for coordinating the submission of P. vivax positive blood samples to the Wadsworth Center (New York Department of Health), for subsequent inclusion in this study.
Disclaimers: The findings and conclusions of this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
Footnotes
Supplementary data related to this article can be found at https://doi.org/10.1016/j.lana.2025.101159.
Contributor Information
Joel L.N. Barratt, Email: jbarratt@cdc.gov.
Brian H. Raphael, Email: elx9@cdc.gov.
Appendix A. Supplementary data
Specimen metadata, genetic distance matrices, haplotype datasheets and other relevant information. Specimen metadata found in Tab A are color coded, where samples of the same color are either repeats of the same sample or are samples linked to the same outbreak cluster; see the notes column for precise details. Positive and negative control samples are shaded gray. For interpretation of other color coding in Tab A refer to the text within colored cells.
Supplementary Methods and Results.
Contains a full description of the AmpliSeq panel.
CYCLONE reference files, including two fasta files, and a BED reference file. Instructions on the use of these files can be found in the Supplementary notes from Barratt et al.17
CYCLONE reference files, including two fasta files, and a BED reference file. Instructions on the use of these files can be found in the Supplementary notes from Barratt et al.17
CYCLONE reference files, including two fasta files, and a BED reference file. Instructions on the use of these files can be found in the Supplementary notes from Barratt et al.17
A PDF file of the tree shown in Figure with the labels on each of the branch tips. Labels include the sample name and reported travel histories for case-patients where available.
References
- 1.WHO . World Health Organization; 2022. World malaria report 2022. [Google Scholar]
- 2.Flannery E.L., Kangwanrangsan N., Chuenchob V., et al. Plasmodium vivax latent liver infection is characterized by persistent hypnozoites, hypnozoite-derived schizonts, and time-dependent efficacy of primaquine. Mol Ther Methods Clin Dev. 2022;26:427–440. doi: 10.1016/j.omtm.2022.07.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Dayananda K.K., Achur R.N., Gowda D.C. Epidemiology, drug resistance, and pathophysiology of Plasmodium vivax malaria. J Vector Borne Dis. 2018;55(1):1–8. doi: 10.4103/0972-9062.234620. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.WHO . World Health Organization; 2023. World malaria report 2023. [Google Scholar]
- 5.Gartner V., Redelings B.D., Gaither C., et al. Genomic insights into Plasmodium vivax population structure and diversity in central Africa. Malar J. 2024;23(1):27. doi: 10.1186/s12936-024-04852-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.CDC The history of malaria, an ancient disease. 2018. https://stacks.cdc.gov/view/cdc/135582
- 7.CDC About mosquitoes in the United States. 2024. https://www.cdc.gov/mosquitoes/about/about-mosquitoes-in-the-united-states.html
- 8.Malecki J.M., Kumar S., Johnson B.F., et al. Local transmission of Plasmodium vivax malaria--palm Beach county, Florida, 2003. MMWR Morb Mortal Wkly Rep. 2003;52(38):908–911. [PubMed] [Google Scholar]
- 9.Blackburn D., Drennon M., Broussard K., et al. Outbreak of locally acquired mosquito-transmitted (autochthonous) malaria–Florida and Texas, may-July 2023. MMWR Morb Mortal Wkly Rep. 2023;72(36):973–978. doi: 10.15585/mmwr.mm7236a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Imwong M., Snounou G., Pukrittayakamee S., et al. Relapses of Plasmodium vivax infection usually result from activation of heterologous hypnozoites. J Infect Dis. 2007;195(7):927–933. doi: 10.1086/512241. [DOI] [PubMed] [Google Scholar]
- 11.Chenet S.M., Schneider K.A., Villegas L., Escalante A.A. Local population structure of Plasmodium: impact on malaria control and elimination. Malar J. 2012;11:412. doi: 10.1186/1475-2875-11-412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Joy D.A., Gonzalez-Ceron L., Carlton J.M., et al. Local adaptation and vector-mediated population structure in Plasmodium vivax malaria. Mol Biol Evol. 2008;25(6):1245–1252. doi: 10.1093/molbev/msn073. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Baniecki M.L., Faust A.L., Schaffner S.F., et al. Development of a single nucleotide polymorphism barcode to genotype Plasmodium vivax infections. PLoS Negl Trop Dis. 2015;9(3) doi: 10.1371/journal.pntd.0003539. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kattenberg J.H., Nguyen H.V., Nguyen H.L., et al. Novel highly-multiplexed AmpliSeq targeted assay for Plasmodium vivax genetic surveillance use cases at multiple geographical scales. Front Cell Infect Microbiol. 2022;12 doi: 10.3389/fcimb.2022.953187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Siegel S.V., Trimarsanto H., Amato R., et al. Lineage-informative microhaplotypes for recurrence classification and spatio-temporal surveillance of Plasmodium vivax malaria parasites. Nat Commun. 2024;15:6757. doi: 10.1038/s41467-024-51015-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Trimarsanto H., Amato R., Pearson R.D., et al. A molecular barcode and web-based data analysis tool to identify imported Plasmodium vivax malaria. Commun Biol. 2022;5(1):1411. doi: 10.1038/s42003-022-04352-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Barratt J., Houghton K., Richins T., et al. Investigation of US Cyclospora cayetanensis outbreaks in 2019 and evaluation of an improved Cyclospora genotyping system against 2019 cyclosporiasis outbreak clusters. Epidemiol Infect. 2021;149 doi: 10.1017/S0950268821002090. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Barratt J.L.N., Shen J., Houghton K., et al. Cyclospora cayetanensis comprises at least 3 species that cause human cyclosporiasis. Parasitology. 2023;150(3):269–285. doi: 10.1017/S003118202200172X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Jacobson D., Zheng Y., Plucinski M.M., Qvarnstrom Y., Barratt J.L.N. Evaluation of various distance computation methods for construction of haplotype-based phylogenies from large MLST datasets. Mol Phylogenet Evol. 2022;177 doi: 10.1016/j.ympev.2022.107608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.MalariaGen, Adam I., Alam M.S., et al. An open dataset of Plasmodium vivax genome variation in 1,895 worldwide samples. Wellcome Open Res. 2022;7:136. doi: 10.12688/wellcomeopenres.17795.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ward J.H. Hierarchical grouping to optimize an objective function. J Am Stat Assoc. 1963;58:236–244. [Google Scholar]
- 22.Yu G. Using ggtree to visualize data on tree-like structures. Curr Protoc Bioinformatics. 2020;69(1):e96. doi: 10.1002/cpbi.96. [DOI] [PubMed] [Google Scholar]
- 23.Jacobson D.K., Low R., Plucinski M.M., Barratt J.L.N. An improved framework for detecting discrete epidemiologically meaningful partitions in hierarchically clustered genetic data. Bioinform Adv. 2023;3(1) doi: 10.1093/bioadv/vbad118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Markin A., Wagle S., Grover S., Vincent Baker A.L., Eulenstein O., Anderson T.K. PARNAS: objectively selecting the most representative taxa on a phylogeny. Syst Biol. 2023;72(5):1052–1063. doi: 10.1093/sysbio/syad028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Langmead B., Salzberg S.L. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9(4):357–359. doi: 10.1038/nmeth.1923. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Li H., Handsaker B., Wysoker A., et al. The sequence alignment/map format and SAMtools. Bioinformatics. 2009;25(16):2078–2079. doi: 10.1093/bioinformatics/btp352. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Brouard J.S., Bissonnette N. Variant calling from RNA-seq data using the GATK joint genotyping workflow. Methods Mol Biol. 2022;2493:205–233. doi: 10.1007/978-1-0716-2293-3_13. [DOI] [PubMed] [Google Scholar]
- 28.Muneer A., Adapa S.R., Silbert S., et al. Autochthonous Plasmodium vivax infections, Florida, USA, 2023. Emerg Infect Dis. 2024;30(6):1214–1217. doi: 10.3201/eid3006.240336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Barratt J.L.N., Plucinski M.M. Epidemiologic utility of a framework for partition number selection when dissecting hierarchically clustered genetic data evaluated on the intestinal parasite Cyclospora cayetanensis. Am J Epidemiol. 2023;192(5):772–781. doi: 10.1093/aje/kwad006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ibrahim A., Manko E., Dombrowski J.G., et al. Population-based genomic study of Plasmodium vivax malaria in seven Brazilian states and across South America. Lancet Reg Health Am. 2023;18 doi: 10.1016/j.lana.2022.100420. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Mitchell C.L., Kennar A., Vasquez Y., et al. Notes from the field: increases in imported malaria cases–three southern U.S. Border jurisdictions, 2023. MMWR Morb Mortal Wkly Rep. 2024;73(18):417–419. doi: 10.15585/mmwr.mm7318a2. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Specimen metadata, genetic distance matrices, haplotype datasheets and other relevant information. Specimen metadata found in Tab A are color coded, where samples of the same color are either repeats of the same sample or are samples linked to the same outbreak cluster; see the notes column for precise details. Positive and negative control samples are shaded gray. For interpretation of other color coding in Tab A refer to the text within colored cells.
Supplementary Methods and Results.
Contains a full description of the AmpliSeq panel.
CYCLONE reference files, including two fasta files, and a BED reference file. Instructions on the use of these files can be found in the Supplementary notes from Barratt et al.17
CYCLONE reference files, including two fasta files, and a BED reference file. Instructions on the use of these files can be found in the Supplementary notes from Barratt et al.17
CYCLONE reference files, including two fasta files, and a BED reference file. Instructions on the use of these files can be found in the Supplementary notes from Barratt et al.17
A PDF file of the tree shown in Figure with the labels on each of the branch tips. Labels include the sample name and reported travel histories for case-patients where available.