https://orcid.org/0000-0002-5171-819X
Figures
Abstract
Sabah, Malaysia, has amongst the highest burden of human Plasmodium knowlesi infection in the world, associated with increasing encroachment on the parasite’s macaque host habitat. However, the genomic make-up of P. knowlesi in Sabah was previously poorly understood. To inform on local patterns of transmission and putative adaptive drivers, we conduct population-level genetic analyses of P. knowlesi human infections using 52 new whole genomes from Sabah, Malaysia, in combination with publicly available data. We identify the emergence of distinct geographical subpopulations within the macaque-associated clusters using identity-by-descent-based connectivity analysis. Secondly, we report on introgression events between the clusters, which may be linked to differentiation of the subpopulations, and that overlap genes critical for survival in human and mosquito hosts. Using village-level locations from P. knowlesi infections, we also identify associations between several introgressed regions and both intact forest perimeter-area ratio and mosquito vector habitat suitability. Our findings provide further evidence of the complex role of changing ecosystems and sympatric macaque hosts in Malaysia driving distinct genetic changes seen in P. knowlesi populations. Future expanded analyses of evolving P. knowlesi genetics and environmental drivers of transmission will be important to guide public health surveillance and control strategies.
Author summary
The zoonotic P. knowlesi parasite is an emerging, yet understudied, cause of malaria in Southeast Asia. Sabah, Malaysia, has amongst the highest burden of human P. knowlesi infection in the world, however, the region is currently understudied. We produced a collection of high-quality P. knowlesi genomes from Sabah, and in combination with publicly available data, performed an extensive population genetics analysis. Our work contributes novel insights for Plasmodium knowlesi population genetics and genetic epidemiology.
Citation: Westaway JAF, Diez Benavente E, Auburn S, Kucharski M, Aranciaga N, Nayak S, et al. (2025) Genomic epidemiology of Plasmodium knowlesi reveals putative genetic drivers of adaptation in Malaysia. PLoS Negl Trop Dis 19(3): e0012885. https://doi.org/10.1371/journal.pntd.0012885
Editor: Nadira D. Karunaweera,, University of Colombo Faculty of Medicine, SRI LANKA
Received: April 30, 2024; Accepted: February 3, 2025; Published: March 12, 2025
Copyright: © 2025 Westaway et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Genomic data produced as part of this work are available at the Sequence Read archive (SRA) of the National Center for Biotechnology Information (NCBI) under the BioProject ID PRJNA1066389 and all bioinformatic and analytical scripts are available at https://github.com/JacobAFW/Pk_Malaysian_Population_Genetics. Publicly available data used as part of our analyses can be found at https://doi.org/10.1038/s41598-019-46398-z and https://doi.org/10.3201/eid2608.190864
Funding: Sample collection and sample processing were supported by the Ministry of Health, Malaysia (grant number BP00500420 and grant number BP00500/117/1002 to GSR); the Australian National Health and Medical Research Council (grant numbers 496600, 1037304 and 1045156 to NMA); the US National Institutes of Health (grant numbers R01AI116472-03 to TW and 1R01AI160457-01 to GSR), and the UK Medical Research Council, Natural Environment Research Council, Economic and Social Research Council, and Biotechnology and Biosciences Research Council (grant number G1100796 to CD). Whole genome sequencing was supported by a Singaporean Ministry of Education Grant (grant number MOE2019-T3-1-007 to ZB), and salary support for bioinformatics and analyses through an Australian NHMRC Ideas Grant (grant number APP1188077 to MG). MF (grant number APP5121190) and MJG (grant number 2017436 ) were supported by NHMRC Emerging Leader 2 fellowships; MJG was also supported by the Australian Centre for International Agricultural Research and Indo-Pacific Centre for Health Security, Department of Foreign Affairs and Trade, Australian Government funded ZOOMAL project (grant number LS/2019/116 to MJG). This research has also been funded by the Australian Government through the Partnerships for a Healthy Region (PHR) initiative (RESPOND project, grant number 79233 to MJG). The views expressed in this publication are the author’s alone and are not necessarily the views of the Australian Government. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Zoonotic transmission of the macaque parasite Plasmodium knowlesi has emerged as the most common cause of human malaria in Malaysia and parts of western Indonesia [1–3]. P. knowlesi infections can cause severe, life-threatening malaria, with a case fatality similar to that of P. falciparum in Southeast Asia despite comparatively lower levels of parasitemia [4,5]. The recent increased reporting of P. knowlesi infections in Southeast Asia has been strongly linked with the encroachment of humans on previously intact habitats of their natural macaque reservoir hosts [6]. Zoonotic transmission of P. knowlesi is thought to occur largely in response to increasingly fragmented landscapes as a result of land clearing and associated agricultural activities, with increased exposure in at-risk workers and local populations in endemic areas to both pig-tailed (Macaca nemestrina) and long-tailed (M. fascicularis) macaques, and the Anopheles Leucosphyrus Group mosquito vectors [7,8]. Worryingly, in contrast to the control of other human Plasmodium species, national WHO malaria elimination goals in Southeast Asia are threatened by the inability of public health measures to target macaque host reservoirs for P. knowlesi [2]. Furthermore, conventional prevention measures such as insecticide-treated bed nets used successfully for other Plasmodium species in the region are limited for P. knowlesi zoonotic infections, primarily acquired at the forest-edge during agricultural work activities [9,10].
Insights gained from genomic analyses of human malaria parasites have advanced our understanding of basic disease biology, drug resistance and malaria epidemiology [11]. Large-scale, collaborative efforts to produce publicly available population-level whole genome data for Plasmodium species of interest, have produced over 20,000 P. falciparum [12] and ~1,800 P. vivax [13] genomes. In contrast, P. knowlesi currently has fewer than 200 whole genomes available from a limited geographic distribution [14–18]. Only 16 reported P. knowlesi genomes are described from the state of Sabah in East Malaysia, despite this area representing among the highest reported number of P. knowlesi cases and disease burden globally to date [19].
Previous studies of P. knowlesi population genetics in Malaysia have identified three genetically divergent populations using a combination of whole-genome sequencing [20] and microsatellite genotyping [21]. One of these populations is restricted to Peninsular Malaysia, whilst the other two are found in Malaysian Borneo. The two overlapping clusters in Malaysian Borneo are derived from the separate macaque reservoir hosts: from long-tailed macaques (Macaca fascicularis [Mf]) and pig-tailed macaques (M. nemestrina [Mn]) [22]. We refer to these clusters as Mf (cluster 1), Mn (cluster 2) and Peninsular (cluster 3) throughout this manuscript. Despite these clearly-defined, genetically divergent populations, previous work further identified distinct subpopulations within the different clusters [15], with evidence of recent positive directional selection [20] and large genetic introgression events between the subpopulations linked to mosquito vectors [15]. In this context, introgression refers to the transfer of genetic information from one cluster to another, resulting from hybridisation and repeated backcrossing. This evidence suggests that P. knowlesi population structure is changing, with changes hypothesised to occur as a result of rapidly altering forest and agricultural ecosystems in Malaysia.
To expand our understanding of the evolving population structure of P. knowlesi across Malaysia, we performed whole genome sequencing on 94 new human infections from diverse landscapes across Sabah, East Malaysia [19]. The newly produced data were combined with 108 (100 included in the analysis) publicly available P. knowlesi genomes derived from clinical infections across Malaysia [14,20]. Leveraging the additional isolates from Sabah, our objective was to perform a comprehensive evaluation of P. knowlesi population structure with a dataset that better represents the distribution of symptomatic infections from passive case detection across Malaysia. We combined genomic data with environmental land cover classification data surrounding knowlesi malaria case villages to better explore the relationship between the genomic and ecological features in Sabah associated with the transmission of P. knowlesi populations. These integrated analyses aim to provide insights to assist in the development of future public health interventions and genomic surveillance efforts.
Methods
Ethics statement
The research was performed in accordance with the Declaration of Helsinki and ethics approval was obtained from the medical research ethics committees of the Ministry of Health, Malaysia and Menzies School of Health Research, Australia.
Sample collection and preparation
We used a combination of newly generated P. knowlesi whole genome sequencing data (n = 94) [23] and archived FASTQ files (n = 108) from P. knowlesi-infected patients in Malaysia. Newly processed samples were collected as part of prospective clinical studies conducted through the Infectious Diseases Society Kota Kinabalu Sabah-Menzies School of Health Research collaboration from 2011 to 2016 across multiple hospital sites in Sabah [4,5]. Patients of all ages presenting with microscopy-diagnosed malaria were enrolled following informed consent. Single species, P. knowlesi infections were confirmed through validated PCR (targeting the 18S small-subunit RNA gene) [24,25] and parasitemia quantified by expert research microscopists. These 94 clinical isolates underwent Illumina whole genome, paired-end sequencing (150 bp), with library preparation conducted using the NEBNext Ultra IIDNA Library Prep Kit (from New England BioLabs Inc., Cat No. E7645). The further 108 samples from a broader geographic range within Malaysia were downloaded from the National Center for Biotechnology Information (PRJEB33025, PRJEB23813, PRJEB1405, PRJEB10288 & PRJN294104) [14,20] (Table A in S1 Text).
Read mapping, variant discovery and genotyping
Variants were detected using a modified version of a previously described workflow [26]. Raw reads were processed using FastQC and cutadapt [27] to determine quality, with subsequent filtering and trimming of reads. The Burrows-Wheeler Aligner (bwa) was then used to map reads to the PKA1-H.1 reference genome [28]. BAM pre-processing steps were applied using Picard version 2.26.1 and the Genome Analysis Toolkit (GATK) version 3.8-1-0 [29]. Notably, two steps in the GATK workflow (base recalibration and indel realignment) require a set of high-quality known variants. As recommended by the Broad Institute for non-model organisms without a reference dataset [30], we took a bootstrapping approach, where we passed a subset of 39 samples through a simplified version of the pipeline and applied hard filtering based on quality score distribution (FS<=2, MQ>=59 & QD>=20), and then passed this conservative variant set into the recalibration steps of the pipeline for another round of variant calling.
SNPs and indels were called using a consensus approach applied to outputs from GATK and bcftools variant callers using a modified version of a previously described workflow [31,32]. A consensus approach was taken to improve accuracy and reduce false positives by filtering for the overlap between the two commonly used tools, which use distinct algorithms and heuristics. For GATK, HaplotypeCaller was used to identify potential variants in each sample, with the resulting GVCF files merged using CombineGVCFs, and joint-genotyping performed using GATK’s GenotypeGVCFs. A similar joint-calling approach was implemented with bcftools using the mpileup and call subcommands. A consensus was taken of the resulting VCF files generating a conservative list of high-quality variants. Finally, SNPs and small indels were filtered using GATK’s VariantFiltration using the same thresholds outlined above.
Data filtering
To reduce noise, errors, and avoid bias in statistical estimates from rare variants, further filtering was applied based on clonality (FWS), genotypic missingness and minor allele frequency (MAF), depending on the downstream analysis. For clonality, within-isolate fixation index FWS [33] was calculated on the full dataset (n = 201) using the moimix package (github.com/bahlolab/moimix) and samples with FWS<0.85 were removed from downstream analyses [15]. FWS is a measure of genetic diversity within an isolate, where the genetic variation within individuals is compared to the genetic variation across and entire population. Prior to filtering samples based on FWS, the non-reference allele frequency (NRAF) was also plotted across the genome for individual samples using ggplot2 [34]. Genotypic missingness and MAF were then calculated using PLINK2 [35,36]. SNPs with MAF <5% or genotypic missingness >25% across the population, and samples with >25% genotypic missingness, were filtered from downstream analyses, as well as those SNPs located in hypervariable regions (Table B in S1 Text).
Characterising population structure
To determine overarching population structure, several complementary strategies were employed including neighbour joining analysis based on identity-by-state (IBS), connectivity based on identity-by-descent (IBD) [37], and ADMIXTURE analysis [38]. IBS, a measure of genetic similarity where two alleles at a given locus are identical, was calculated with PLINK and visualised with neighbour-joining trees (NJT) [39] in R using ggplot2 and ggtree [40]. IBD, a measure of genetic similarity where alleles are considered identical if they were inherited from a common ancestor, was calculated with hmmIBD [41], which implements a hidden Markov model to determine sequence segments of shared ancestry. Base R and igraph [42] were used for IBD visualisation at a variety of thresholds (represent the percentage of the genome that is IBD between pairs of samples). To determine the proportions of mixed ancestry, ADMIXTURE was used to implement a maximum likelihood estimation, which was then visualised in R using ggplot2. CV error was calculated prior to ADMIXTURE analysis to identify the optimal K value. As K=3 was deemed optimal, exhibiting a low cross-validation error compared to other K values determined by ADMIXTURE’s cross-validation procedure, and the distribution of samples aligned with the NJT, the K clusters were referred to throughout the manuscript with the previously defined Peninsular- and macaque-associated cluster names (Macaca fascicularis (Mf), Macaca nemestrina (Mn) & Peninsular).
Identifying the presence of introgression events
We performed a bespoke analysis to identify possible genomic regions of introgression. First, we identified the major allele for each cluster at each genomic coordinate. Then using a sliding window approach (10kb windows), we determined the genetic distance for each sample to each cluster. Genetic distance was defined as the proportion of mismatched SNPs per sliding window (10kb) when comparing the called allele in the sample to the major allele for a cluster at each position. The genetic distances were then plotted on a two-dimensional axis, with different clusters along the x and y axis, and two-dimensional kernel density estimations (contours) were calculated for the genomic clusters using ggplot2 (github.com/tidyverse/ggplot2) and MASS (github.com/cran/MASS) packages, and the density contours overlayed on the plot. The points.in.polygon function (github.com/edzer/sp) was used to determine in which contours the windows are spatially located. Windows located within the contours of another cluster whilst outside the major contours of their own were defined as introgressed. To be conservative, candidate windows underwent several filtering steps, including those that appear in>= 5% of the population and the removal of windows that overlap hypervariable regions (Table B in S1 Text).
Exploring links between introgression and environmental land types
We performed subsequent regression analyses to explore whether surrounding village-level environmental land types and predicted vector habitat suitability are associated with P. knowlesi introgression. The primary residential addresses for deidentified P. knowlesi cases for the preceding 3 weeks before health facility presentation was first used to obtain centroid village-level location coordinates cross-checked for accuracy using Google Earth (version 7.3). Selected environmental classification metrics of forest fragmentation (percentage of Landscape – tree cover and Perimeter-Area Ratio – tree cover) within a 5km radius surrounding village locations were then calculated from a composite landscape metrics tool encompassing ESRI 2020 and Sentinel-2 GIS data at 10-metre resolution [43] (Figs A and B in S1 Text). The relative predicted Anopheles Leucosphyrus Complex mosquito vector occurrence surface from Moyes et al. [44] based on boosted regression tree models encompassing mosquito sampling presence/absence data (1999-2014) and environmental covariates indicating habitat suitability was obtained through the malariaAtlas R package [45]. The mosquito vector habitat suitability surface was averaged within a 5x5km grid around the geolocated village sites. Moran’s I [46] was calculated to exclude spatial autocorrelation of the environmental land types and predicted vector habitat suitability at the selected grids. Univariate regression analyses were initially used to assess potential associations between these environmental parameters and the macaque-derived clusters. Tertiles were then generated representing the degree of introgression in samples, with samples categorised as having low, medium, or high introgression. Logistic regression models were subsequently implemented to assess if landscape fragmentation indices or Anopheles Leucosphyrus Complex habit suitability were associated with either the presence of the top ten most frequently introgressed windows (binomial) or the degree of introgression (ordinal). The Akaike Information Criterion (AIC) [47] was compared to determine the optimal model design (scripts available in the attached GitHub repository).
Identification of orthologous antimalarial drug resistance markers
Antimalarial drug resistance markers for P. falciparum and P. vivax (putative) were collated using multiple sources [48,49] and their orthologues in P. knowlesi identified. This includes dihydrofolate reductase (dhfr), dihydropteroate synthase (dhps), chloroquine resistance transporter (crt), multidrug resistance protein 1 (mdr1), multidrug resistance-associated proteins 1 (mrp1), plasmepsin 4 (pm4), kelch 13 (k13), reticulocyte binding protein 1a (rbp1a) and reticulocyte binding protein 1b (rbp1b). P. falciparum and P. vivax orthologues were identified using PlasmoDB’s [50] Orthologue and synteny tool, which is based on the OrthoMCL database, a genome-scale algorithm for grouping orthologous protein sequences [51]. Multiple sequence alignment was then performed on PlasmoDB sequences, comparing the P. knowlesi amino acid sequences against the relevant orthologues in P. falciparum and P. vivax to identify shared mutations between orthologue genes.
Characterising subpopulations within Malaysian Borneo
Once the larger P. knowlesi population structure was characterised, we investigated differences between sub-population clusters within Malaysian Borneo and interrogated the genomes of all relevant samples for signs of differentiation. Samples were subset to the major genomic clusters (Mf and Mn) of Malaysian Borneo, using their distribution on the NJT and major population assigned by ADMIXTURE. Then, both IBS and IBD analyses were repeated on these subsets (Mf and Mn specific subsets) to determine whether subpopulations exist within these larger populations. PLINK was used to calculate the fixation index (FST), a measure of population differentiation due to genetic structure, specifically, the variance of allele frequencies between populations. ggplot2 was used to visualise FST across the genome using a non-overlapping sliding window (1-kb) approach. This allowed the identification of outlier regions, which were annotated to identify potential genes of interest. Further details and scripts for the methods described can be found at https://github.com/JacobAFW/Pk_Malaysian_Population_Genetics.
Results
The 94 newly sequenced P. knowlesi whole genomes all originated from the state of Sabah, encompassing human infections from 11 administrative districts, including 22 infections from Kota Marudu and 14 from Kudat (Table A in S1 Text), collected between May 2011 and February 2016. These genomes had an average of 66,415,402.53 reads per sample, with 30.27% mapping to the PKA1-H.1 reference genome [28]. The average sequencing depth (excluding mitochondria and apicoplast) of these new genomes was 80X (2 to 362X across samples), with 82.5% (18.9 to 97.2% across samples) of the bases in the reference genome covered. The distribution of reads across chromosomes was relatively even, with the mean sequencing depth ranging from 70 to 87X, and the mean percentage of bases covered ranging from 80.3% to 84.0%. The 108 high-quality publicly available P. knowlesi genomes from NCBI were derived predominantly from human infections (with six laboratory strains passaged through macaques) across different districts from both Peninsular Malaysia (n=33) and East Malaysia on the island of Borneo, including the geographically distinct neighbouring state of Sarawak (n=59) in addition to a small number from Sabah (n=16). FWS analysis was performed on 201 samples, with additional filtering based on clonality, missingness, and minor allele frequency, used to generate a subset for other analyses. 52 of the newly sequenced genomes remained after the additional filtering, and another 100 genomes from the publicly available data. The combined 152 P. knowlesi genomes consisted of: 61 from Sabah, 59 from Sarawak, and 32 from Peninsular Malaysia. Joint genotyping initially identified 1,542,627 single nucleotide polymorphisms (SNPs), which after filtering (clonality, missingness and minor allele) resulted in 357,379 SNPs.
Complex P. knowlesi infections in Malaysian Borneo
Given that P. knowlesi parasites are haploid in the blood stage of host infection, the presence of multiple alleles at given loci is indicative of a multiple clone (polyclonal) infection. The within-isolate fixation index (FWS) was used to measure the genetic complexity of all infections (n = 201). The FWS score ranges from 0 to 1, with increasing values reflecting increasing clonality [52]. At a commonly applied threshold of FWS < 0.95, 13.4% (n=27/201) of infections were polyclonal. The highest proportion of polyclonal infections was observed in Sarawak (17.6%, n=13/74), followed by Peninsular Malaysia (12.1%, n=4/33) and Sabah (10.6%, n=10/94). Sarawak had significantly lower FWS than both Sabah (p < 0.05) and Peninsular Malaysia (p < 0.05) (Figs 1A and C in S1 Text).
A: Boxplots depicting the distribution of within-infection diversity (FWS) across three regions (Sabah, Sarawak, and Peninsular). B: Dot plots of the within-isolate non-reference allele frequencies (NRAF) across the genome for six Malaysian P. knowlesi infections ranging from low diversity (FWS=0.99) to high diversity (FWS=0.54) and with varying levels of within-host relatedness.
Non-reference allele frequency (NRAF) plots illustrate a variety of within-host diversity patterns. This includes distinct clones (18.5% [n=5/27]) (e.g., ERR985376 [FWS = 0.93]) and genetically mixed clones (e.g., ERR985395 [Fws = 0.54]). Samples with distinct clones could be the result of either superinfection (multiple mosquito inoculations) or co-transmission (single mosquito inoculation). However, those that are genetically mixed are likely the result of co-transmission events, as given adequate genetic diversity in a population (not inbred), superinfections with highly related clones are unlikely.
P. knowlesi genomes from Sabah belong predominantly within the Mf cluster
Previous genetic studies have described distinct genetic clustering of P. knowlesi into a geographic Peninsular-Malaysia sub-population, and two Malaysian-Borneo macaque-associated subpopulations; M. fascicularis (Mf) and M. nemestrina (Mn) [17,22,21]. We sought to determine the genetic clustering patterns of the Sabah genomes relative to infections from Sarawak and Peninsular Malaysia. Neighbour-joining analysis based on identity-by-state (IBS) was undertaken on the 152 low complexity P. knowlesi genomes from across Malaysia, revealing three clusters (Fig 2A). The newly sequenced P. knowlesi samples originating from Sabah group predominantly within the Mf cluster (82.7%, n=43/52); the remaining (17.3%, n=9/52) infections clustered within the Mn clade, similar to the proportions of the 164 samples from Sabah previously described by Divis et al. 2017 [21] (Mf = 86.6%, Mn = 13.4%) [21]. ADMIXTURE analysis revealed the greatest likelihood of 3 sub-populations amongst the 152 infections (Fig F in S1 Text), confirming the patterns observed with neighbour-joining analysis (Fig 2C).
A: Unrooted neighbour-joining tree based on identity-by-state (IBS) depicting three predominant genomic clusters of P. knowlesi across Malaysia, specifically the Peninsular Malaysia sub-population (Peninsular), and Malaysian-Borneo macaque-associated subpopulations of Macaca fascicularis (Mf) and M. nemestrina (Mn). The new isolates from Sabah are labelled separately (Sabah). B: Map of Malaysia showing the geographic distribution and number of samples, and genomic clusters across Malaysian-Borneo (right) and Peninsular-Malaysia (left). C: Bar plot illustrating the proportionate ancestry to each of 3 (K) subpopulations determined by ADMIXTURE for each sample (bars on x-axis), sectioned by geographic region. The three K populations identified aligned perfectly with the clustering in the NJ tree; K=1 with Mf, K=2 with Mn and K=3 with Peninsular as per the colour-coding. Shapefile made with Natural Earth: https://www.naturalearthdata.com/downloads/50m-cultural-vectors/50m-admin-1-states-provinces/.
Despite substantial genetic divergence between the Mf and Mn clusters (mean FST = 0.2), there is also substantial geographic overlap (Fig 2B) and evidence of shared ancestry (>1% ancestry to two or more groups) amongst 6.6% (n=10/152) of infections (Fig 2C). This observation extends beyond the newly sequenced Sabah samples and to those previously reported in neighbouring Sarawak, with the separate genomic Mf and Mn clusters and several samples of Mf and Mn ancestry being identified in both geographic locations. P. knowlesi infections with shared ancestry originated from the Sarikei and Betong districts in the state of Sarawak and four of the newly sequenced Sabah infections (from Papar, Ranau and Kota Marudu districts). Although several Malaysian-Borneo P. knowlesi infections had evidence of shared ancestry, samples from Peninsular-Malaysia and the district of Kapit in Sarawak (both Mn and Mf) are descendants of single ancestral populations (Fig 2C).
Greater genetic diversity within the Mf than Mn and Peninsular clusters
Since malaria parasites are recombining organisms, neighbour-joining analysis can miss recent connectivity between infections where outcrossing has taken place. To further elucidate the relatedness between isolates, both within and across clusters, we performed identity-by-descent (IBD) analysis on the 152 low complexity infections. In IBD analysis, genomic segments are characterised as identical by descent in pairwise comparisons when identical nucleotide sequences have been inherited from a common ancestor. The Mf cluster had the highest genetic diversity, with a median IBD of 7.0%, and as such, we see most of the connectivity break down at a relatively low threshold of 10% (Fig 3). In contrast, the Mn (median IBD = 0.5) and Peninsular (median IBD = 0.3) clusters maintain tight networks at 25% IBD, reflective of more recent common ancestry and a greater number of shared haplotypes, and in turn, lower transmission intensity (Fig 3). An Mf isolate (PK_SB_DNA_028 – Papar, Sabah) also maintains connectivity with the Mn cluster at an IBD threshold of 10%, with the regions of IBD between the Mf isolate and the Mn isolates consistent across pairwise comparisons (Fig G in S1 Text). The Mf isolate was collected in Papar (Sabah), and the Mn isolates from Kapit, Sarikei and Betong (Sarawak). To confirm the high IBD values in Mn and Peninsular clusters were not inflated by the SNPs used (i.e., being biased by strong population structure), we trialled multiple filtering combinations and re-calculated the median IBD values for comparison, confirming our initial findings (Table C in S1 Text).
Each circle reflects an infection, colour-coded by genomic clustering group, and the number of lines between infections reflects relatedness (more lines reflect greater relatedness) at the given connectivity threshold of minimum IBD. Where two circles are not connected by a line, the estimated IBD between those infections was below the given threshold. The three samples from Peninsular Malaysia with >95% IBD represent laboratory-based strains from the 1960s that have been passaged through macaques (SRR2222335, SRR2225467 & SRR3135172).
Greater relatedness within state-level P. knowlesi subpopulations
It was hypothesised that P. knowlesi clinical infections derived from the separate states of Sabah and Sarawak in Malaysian Borneo are likely to have distinct genetic ancestry due to factors such as differences in the primary Anopheles Leucosphyrus Group mosquito vector species and other large-scale environmental features that may have restricted historical gene flow [53]. To test this, we leveraged the newly sequenced genomes to perform additional analyses on a subset of the data comprising isolates from Sabah and Sarawak. We examined the potential impact of geographical regions on population structure and genetic relatedness within each of the separate Mf and Mn clusters, performing IBD analyses on the clusters separately. IBD analyses of Mf and Mn subsets suggest that most samples have greater connectivity within their respective states (Fig 4). Two Sabah samples within the Mf cluster had a high degree of connectivity with Sarawak samples (Fig 4A). The samples (PK_SB_DNA_028 and PK_SB_DNA_053) are P. knowlesi infections collected from residents of the Papar and Kudat districts in Sabah.
A & B: Identity-by-descent (IBD)-based cluster network illustrating the distant relatedness for samples within Mf (A) and Mn (B) clusters collected in two adjacent states; Sabah and Sarawak, at different cut-offs for the proportion of IBD in a paired comparison. C: Map of East Malaysia on the island of Borneo with colours representing the two states being compared in the IBD analysis. Shapefile made with Natural Earth: https://www.naturalearthdata.com/downloads/50m-cultural-vectors/50m-admin-1-states-provinces/.
Population differentiation within Mf and Mn clusters of geographic subpopulations
Given that sampling sites of the P. knowlesi geographic subpopulations are isolated by several hundred kilometres, with spatially heterogenous environmental pressures, we performed genome-wide scans for differentiation between the geographic subpopulations within Mf and Mn subsets. Genome-wide scans within Mf highlighted several regions of significant differentiation across the genome, appearing as peaks of multiple tightly clustered windows of high FST against a background of low differentiation (mean FST = 0.007, Fig 5A). The most notable peaks within the Mf cluster were observed on chromosomes 8, 11 and 12. The peak on chromosome 8 covers a region containing the gene encoding for the oocyte capsule protein, with a complete list of genes found in the peak available in Tables H and I in S1 Text. Unfortunately, due to substantial noise, it was not possible to appropriately identify peaks for the Mn cluster (mean FST = 0.036, Fig 5B).
Genome-wide scans of differentiation between Sabah and Sarawak subpopulations within the (A) Mf and (B) Mn clusters using the between-population fixation index (FST). Only the Mf cluster shows clear differences in diversity with peaks of differentiation clear at several chromosomes (most notably on chromosomes 8, 11 and 12), whilst the Mn cluster has substantial ‘noise’ across the genome, with high levels of differentiation across the genome.
Substantial evidence for introgression between Mn and Mf clusters
Previous studies have described the occurrence of chromosomal-segment exchanges between the Mn and Mf subpopulations, suggesting that they are not genetically isolated [15]. We therefore sought evidence for introgression events in our large collective cohort, and specifically, in the previously underrepresented state of Sabah. Comparisons of genetic distance between 10kb sliding windows in individual P. knowlesi-infected samples and different clusters reveal evidence of substantial genetic exchanges across genomic clusters, chromosomes, and geographical regions (Fig 6 and Table D in S1 Text). The degree of introgression, represented by the number of introgressed windows identified in a sample, also varied between all of the above-mentioned features.
A: Dot and contours plot describing potential introgression events within an Mf sample, where x and y axes represent the genetic distance of the sample to the Mf and Mn clusters, respectively. Genetic distance is the proportion of mismatched SNPs per sliding window (10kb) when comparing the called allele in the sample to the major allele for a cluster at each position. The contours represent the density of genetic distances for the three clusters. B: Dot and contours plot of the same sample above, subset to those windows deemed to be introgressed from the Mn cluster. Possible introgression events are sliding windows that fall outside the major contours of the samples own cluster and within the major contours of another, representing greater similarity in genetic distance to the other cluster. C: Unrooted neighbour-joining tree based on identity-by-state (IBS) of window 1504 on chromosome 08 (950000-959999) and overlapping the PKA1H_080026000 gene (encodes the oocyst capsule protein). The Mf samples/branches clustering within the Mn branches (depicted by asterisk) provides further evidence that introgression of this window has occurred in these samples. D: SNP barcode plot of window 1504 on chromosome 08 (950000-959999) showing greater genetic similarity between several Mf samples (depicted by asterisk) and the Mn cluster, where the colours reflect those in the legend on panel C, and the alpha represents the allele.
Of the 152 individual P. knowlesi samples analysed, 71.1% (111/152) had introgressed windows (10kb). Given the complexity of distinguishing single biological introgression events, due to potential ambiguity when grouping adjacent windows with varying start and stop positions across samples (Fig H in S1 Text), we define the 10kb windows identified in our analysis as introgressed windows without making assumptions about the underlying biological events. Approximately 29.5% (n=46/152) of samples demonstrated a high degree (>5 10kb windows) of introgression. Within the subset of newly generated genomes from Sabah, 82.7% (n=43/52) samples had two or more introgressed windows, including 20 with >5 windows. The Mf cluster had a higher median number of introgressed windows per sample (median 5, IQR ) compared to Mn (median 1, IQR
). Introgressed windows across ten chromosomes for Mf and six chromosomes for Mn. For the Mf cluster, chromosomes 8 (n = 35) and 11 (n = 21) had the greatest number of introgressed windows, with several windows on chromosome 8 overlapping the large peak observed in the Fst analysis (Fig 5). For Mn, all six chromosomes contained a single window.
The district of Betong in Sarawak had the highest median number of introgressed windows per individual P. knowlesi sample (median 29, IQR ) followed by the district of Papar in Sabah (median 19, IQR
). 85% (n =12/14) of samples from Betong and 50% (n=1/2) from Papar had high levels of introgression, with all but one Betong sample from the Mf cluster. The Mf isolate from Papar with high levels of introgression (PK_SB_DNA_028), had the greatest number of windows overall (n = 37), followed by ten Mf samples from Betong that had greater than 20 introgressed windows (Table D in S1 Text). This same sample, (PK_SB_DNA_028) is also the Mf isolate that shared a higher degree of IBD (10%) with the Mn cluster relative to its own cluster (Fig 3), suggesting that introgression events may be a contributing factor to shared regions of IBD between samples (Fig G in S1 Text). Furthermore, this mechanism also explains the IBD-based connectivity between the two Sabah samples (PK_SB_DNA_028 and PK_SB_DNA_053) and Sarawak samples in the cluster-specific IBD analysis (Fig 4A), as both samples also exhibit substantial patterns of introgression, in a similar pattern (the same or proximal windows) to that seen in the samples from Sarawak (Table G in S1 Text).
Several candidate introgressed regions identified in the Mf clusters overlap putative genes involved in host interactions. Within Sabah, the most common candidate window (window 1504, chromosome 11: 2080000 - 2089999), observed in both Sabah (n = 17) and Sarawak (n = 7) isolates, overlaps a gene encoding for the parasite DNA repair protein RAD50 (PKA1H_110050600), which may aid survival within the host [54]. Focusing on the top 10 most abundant windows in both Sabah and Sarawak, several other genes encoding for proteins essential for survival or invasion in the human, macaque or mosquito hosts were also identified to overlap candidate windows (Table 1). Amongst windows more prevalent in samples from Sarawak were several genes that encode for mosquito-related proteins. This includes the oocyst capsule protein (PKA1H_080026000), CPW-WPC family protein (PKA1H_080026200) and the microneme-associated antigen (PKA1H_080031400). The inclusion of the new Sabah isolates expands the distribution of the introgression event associated with the oocyst-expressed cap380 gene, previously only observed in Betong, Sarawak [15,18]. The oocyst capsule protein is essential for the maturation of ookinete into oocyst in P. berghei and is assumed to assist in immune evasion in mosquito hosts [55]. The CPW-WPC proteins are zygote/ookinete stage-specific surface proteins and appear to be involved in mosquito-stage parasite development [56] and micronemes are critical for host-erythrocyte invasion [57]. The number of candidate regions in Mn was minimal (n = 6), with even fewer involved in host interactions. The exception being window 807 on chromosome 08 (760000-769999), overlapping the PKA1H_080021900 gene, which is essential for erythrocyte invasion in P. falciparum [58]. It should be noted that although the biology of these putative genes is well understood in other human-only Plasmodium species, they may not directly translate to the biologically and genetically distinct P. knowlesi.
Ecological pressures driving introgression
In order to evaluate whether the cluster distribution or the introgression events are associated with ecological changes that might impact either macaque host or vector adaptations, we collated satellite-based surrounding forest fragmentation data and mosquito vector habitat suitability for 37 P. knowlesi samples in Sabah where village locations could be obtained (Figs A and B in S1 Text). These samples included 29 (78.4%) with two or more windows where introgression was observed, and 13 (35.1%) with high introgression (>5 windows).
Firstly, we performed univariate regression analyses of the P. knowlesi genomic clusters against proportional forest cover, intact forest perimeter-area ratio and Anopheles Leucosphyrus Complex mosquito vector habitat suitability metrics, with no statistically significant associations. Secondly, univariate regression analyses (optimal model as determined by AIC comparisons) suggested a limited relationship between two introgression windows, 859 (chromosome 08: 1280000 – 1289999) and 1236 (chromosome 10: 920000 – 929999) and the intact forest perimeter-area ratio and mosquito vector habitat suitability, respectively (Tables E and F in S1 Text). The introgression window 859, which contains no identifiable genes on PlasmoDB and was identified in three Sabah and ten Sarawak isolates, was positively associated with intact forest perimeter-area ratio (ꭓ2 = 6, df = 1, p = 0.02, r2 = 0.69). The introgression window 1236, which was identified in eight Sabah and two Sarawak isolates, was negatively associated with the predicted mosquito vector habitat suitability (χ2 = 8.17, df = 1, p < 0.01, r2 = 0.71). Putative genes overlapping this region include two encoding for unknown proteins, one encoding for ras-related protein Rab-1B and another for orotate phosphoribosyltransferase (Table 1).
Investigation of antimalarial drug resistance candidates in P. knowlesi orthologues
The presence of antimalarial drug resistance determinants in P. knowlesi infections could be considered a surrogate marker of human-human transmission given the absence of drug pressure in the macaque hosts and fitness costs that are often associated with resistance-conferring alleles [59]. We therefore investigated the prevalence of non-synonymous variants in P. knowlesi orthologues of genes that have previously been associated with P. falciparum and P. vivax resistance to antimalarial drugs [48,49]. Within low complexity infections (n=152), six non-synonymous variants were detected within the P. knowlesi orthologue of pvdhps (PKA1H_140035100) (Table 2), which may be linked to sulphadoxine resistance, although very few studies associate genotype and phenotype [49]. The most common variants occurred at codon Y308H (58.4%) and K66E (11.9%). A G422S variant was present in 4.3% of samples overall, although was found exclusively in 25% of isolates from Peninsular Malaysia. Similar to previous work [59], 13 non-synonymous mutations were also detected within the P. knowlesi orthologue to pvdhfr (PKA1H_050015200) (Table 2). This includes 17 samples with greater than one mutation, and two samples with three mutations. Dihydrofolate-reductase mutations, associated with resistance to pyrimethamine, arise readily in both P. falciparum [60] and P. vivax [61,62]. The most common mutations were at codon N272S (97.9%) and E262D (22.7%), with N272S occurring 3 amino acid positions from the P. vivax orthologue. Six mutations were also observed exclusively in isolates from Peninsular Malaysia, with a mean frequency of 10.6%. Lastly, several non-synonymous mutations were also observed in the PKA1H_140054000 gene (Table 2), resulting in several amino acid changes near those observed in P. vivax, and associated with the multidrug resistance protein 1 (pvmrp1). This includes seven amino acid substitutions that occur <3 amino acids from non-synonymous mutations present in pvmrp1, which has a potential yet unconfirmed role in primaquine failure in P. vivax liver-stage infection relapse. However, the biology and significance of these SNPs in these putative genes may not translate directly from P. falciparum and P. vivax.
Drug resistance orthologues from P. vivax (putative) and P. falciparum (none identified from P. falciparum) identified in this P. knowlesi dataset. +Designates the number of amino acid positions the identified P. knowlesi mutation is from the corresponding P. vivax orthologue mutation position (proximal mutations may retain the potential to cause similar downstream effects). AA: amino acid change; Pk: P. knowlesi. Pv: P. vivax.
Discussion
This study expands current understanding of P. knowlesi population genetics by incorporating additional whole genomes from Sabah, a key transmission area in Malaysia. We identified distinct geographical subpopulations within Mf- and Mn-associated clusters, with evidence of introgression between these clusters potentially driving differentiation. Preliminary ecological-genomic analysis suggests possible associations between genomic patterns and environmental features affecting host or vector adaptation. Additionally, we detected non-synonymous mutations in antimalarial drug resistance-related orthologous genes arising de novo given the zoonotic transmission mode and lack of drug selection pressure [63,64].
Within-host Plasmodium genetic diversity reflects transmission intensity, with superinfections arising from multiple mosquito bites or co-transmission of related parasite strains in a single bite [11,65,66]. The prevalence of human polyclonal P. knowlesi infections were lower than in P. falciparum or P. vivax endemic regions [67,68], although are likely higher in natural macaque hosts [69]. The zoonotic nature of P. knowlesi complicates infection complexity, with multiple underlying parasite, host and epidemiological factors potentially influencing the establishment of successful erythrocytic replication of multiple inoculated P. knowlesi strains within humans. These factors include specific parasite proteins involved in human red blood cell invasion including PkDBPαII and PkNBPXa [70], transmission intensity and the relationship with parasite genetic diversity in macaque hosts, and the impact of land use change on mosquito distribution and host biting preferences [71,72]. Inter-infection diversity can also be exacerbated by recombination between genetically distinct parasites within the mosquito (coinfections) [11]. Being a natural reservoir for multiple zoonotic Plasmodium species, macaques have been shown to be co-infected with up to five simian Plasmodium species [69] and multiple P. knowlesi clones [73,74]. Although it cannot be confirmed with the methods used here, the isolates with distinct clones could represent superinfections. While there is no evidence of sustained human-to-human transmission to date [75], high-risk groups like forestry and plantation workers face greater exposure to infected vectors and reservoir hosts [9,10,76]. The five individuals harbouring multiple distinct clones could belong to these at-risk groups, warranting further research with integrated epidemiological and genomic datasets.
Both neighbour-joining and IBD-based cluster analyses identified the three major known P. knowlesi genomic clusters in Malaysia. The majority of new isolates from Sabah belong to the Mf cluster, aligning with reports of higher prevalence of Mf-derived infections and the restricted habitat of M. nemestrina in intact forests [22,44]. The low median IBD in the Mf cluster suggests high transmission intensity and genetic diversity, typical of endemic Plasmodium populations with minimal inbreeding. In contrast, the higher IBD values in the Mn cluster suggests greater parasite relatedness and possible inbreeding, however, as the median IBD reduces substantially when down sampling to the Mn cluster, these values may be skewed by population structure [14]. The broad ecosystem range and adaptability of M. fascicularis [44] may contribute to the Mf cluster’s higher genetic diversity, which could hinder malaria control efforts by enhancing the parasite’s ability to adapt to environmental changes and broaden efficient zoonotic transmission scenarios.
Deforestation and agricultural expansion have altered macaque and Anopheles habitats, likely driving recent genetic exchanges in human infections [14,72,76,77]. Regression analyses revealed significant associations between two introgressed windows and both forest fragmentation (perimeter-area ratio) and habitat suitability of the Anopheles Leucosphyrus Complex mosquito vector, broadly supporting this hypothesis at a population-level. These genomic regions contain putative genes critical for parasite survival and transmission, such as the microneme-associated antigen, which facilitates erythrocyte entry [78], and the oocyst-expressed cap380 gene, essential for vector-stage transmission, previously identified in Betong, Sarawak [14,55]. Apicortin protein, vital for cytoskeletal stability, replication, and host erythrocyte invasion in P. falciparum and P. vivax, was also identified [79,80]. The presence of several human- and vector-related genes in introgressed windows suggests strong selective pressure from both hosts. However, while these genes are well-characterized in other Plasmodium species, their functions may not directly translate to P. knowlesi.
Introgression events occurring across large geographic distances suggest independent occurrences driven by similar environmental drivers, like deforestation and shifting vector populations. This is supported by the non-overlapping introgressed windows among isolates from different geographic regions (Fig I in S1 Text). However, this integrated genomic and spatial analysis is limited by the small subset of isolates and landscape metrics used [81], as well as the lack of temporal alignment between P. knowlesi isolate collection and environmental data, especially when one considers ongoing deforestation, reforestation and land use change in Sabah. Future work should involve larger sample sizes and a systematic approach to landscape classification, including accounting for temporal land-use changes.
Sampling of P. knowlesi infections in Malaysian Borneo occurred across two large geographical areas. While macaque host infection prevalence and transmission intensity at a troop level is the likely primary driver of P. knowlesi population structure, environmental factors likely influence structure across heterogenous landscapes. To assess geographical impacts, we analysed the Mf and Mn clusters separately, comparing Sabah and Sarawak subpopulations. As expected, samples collected in closer proximity showed higher relatedness, albeit less so than across the three major genomic clusters. Notably, within Mf two Sabah P. knowlesi isolates with a high degree of introgression clustered with Sarawak samples. One individual was from a village in Kudat but has a history of recent travel to Hutan Long Pasia in Sipitang district for work, which is located close to the Sarawak border. However, the other individual has no history of recent travel, suggesting this could be the result of independent introgression events arising across regions or the small possibility of an onwards human transmission event.
The stronger relatedness between geographically proximal samples suggests ecological pressures, alongside macaque hosts, influence P. knowlesi genomes. FST analysis of the Mf cluster identified several vector-related genes, also found in introgression analysis, including the oocyst-expressed cap380 gene on chromosome 08. This finding may suggest regional differences in the mosquito vector species within the Anopheles Leucosphyrus Complex [82] contributes to P. knowlesi subpopulation variation. In Sabah, human land use change has altered vector behavior, breeding sites, and biting preferences in the primary vector A. balabacensis [83]. Future studies may benefit from using cluster-specific reference genomes for FST analysis [84], particularly for the Mn cluster.
As P. knowlesi transmission appears exclusively zoonotic [2], and therefore without drug-pressure, resistance mutations are unlikely to arise. The dhfr and dhps mutations may be associated with resistance to pyrimethamine and sulphadoxine, previously used to treat P. falciparum in Malaysia [85]. However, artemisinin-combination therapy is now the recommended treatment for uncomplicated malaria in Malaysia, including P. knowlesi, eliminating the potential for ongoing sulphadoxine-pyrimethamine selection pressure [86,87,88]. We have also previously showed that dhfr mutations in P. knowlesi are unlikely to be due to sulphadoxine-pyrimethamine selection pressure due to not occurring in the drug binding domain [59], and no dhps mutations associated with resistance have been identified [17]. The absence of proven natural human-to-human transmission and sulphadoxine-pyrimethamine use for P. knowlesi infections, suggests that these mutations likely reflect the polymorphic nature of these genes rather than drug selection pressure. The highly prevalent N272S mutation in dhfr appears fixed in the population, with the reference allele (A) found only in older lab-adapted lines, originally collected in the 1960’s, and the A1.H.1 strain, while the alternate allele (G) dominates recent populations and the PKNH reference genome [89,90]. Lastly, although mrp1 variants have been reported [16], the functional impact of the SNPs observed here remains unclear, as the biology may not directly translate from P. falciparum or P. vivax.
The addition of 52 high-quality P. knowlesi genomes from Sabah, Malaysia enhances our understanding of this unique parasite’s evolving genomic landscape. We identify polyclonal infections and describe novel regional P. knowlesi within-cluster subpopulations, likely driven by introgression between the Mf- and Mn-associated clusters. These genomic introgression events in turn may reflect ecological influences from host or vector adaptations. Human encroachment on ecosystems through anthropogenic deforestation and agriculture appears to align with these genetic changes. Additionally, non-synonymous mutations were found in dhps, dhfr and mrp1 putative drug-resistant genes. Insights from P. falciparum and P. vivax highlight the importance of expanding and adapting integrated genetic, epidemiological and environmental surveillance efforts to address the zoonotic context of P. knowlesi when developing future public health control strategies.
Supporting information
S1 Text. Masked genomic regions and supplementary outputs.
https://doi.org/10.1371/journal.pntd.0012885.s001
(DOCX)
Acknowledgments
We thank the study participants, and the research team at the Infectious Disease Society Kota Kinabalu Sabah including Sitti Saimah binti Sakam, Azielia Elastiqah binti Salamth and Mohd Rizan Osman. We thank the Director-General, Ministry of Health, Malaysia, for permission to publish this manuscript. We thank Dr Freya Shearer and Dr David Duncan from the University of Melbourne for their consultation on specific analyses.
References
- 1. Cooper D, Rajahram G, William T, Jelip J, Mohammad R, Benedict J. Plasmodium knowlesi malaria in Sabah, Malaysia, 2015–2017: Ongoing increase in incidence despite near-elimination of the human-only Plasmodium species. Clinical Infectious Diseases. 2020;70(3):361–7.
- 2. Fornace KM, Drakeley CJ, Lindblade KA, Jelip J, Ahmed K. Zoonotic malaria requires new policy approaches to malaria elimination. Nat Commun. 2023;14(1):5750. pmid:37717079
- 3. Lubis IND, Wijaya H, Lubis M, Lubis CP, Divis PCS, Beshir KB, et al. Contribution of Plasmodium knowlesi to Multispecies Human Malaria Infections in North Sumatera, Indonesia. J Infect Dis. 2017;215(7):1148–55. pmid:28201638
- 4. Barber BE, William T, Grigg MJ, Menon J, Auburn S, Marfurt J, et al. A prospective comparative study of knowlesi, falciparum, and vivax malaria in Sabah, Malaysia: high proportion with severe disease from Plasmodium knowlesi and Plasmodium vivax but no mortality with early referral and artesunate therapy. Clin Infect Dis. 2013;56(3):383–97. pmid:23087389
- 5. Grigg MJ, William T, Barber BE, Rajahram GS, Menon J, Schimann E, et al. Age-Related Clinical Spectrum of Plasmodium knowlesi Malaria and Predictors of Severity. Clin Infect Dis. 2018;67(3):350–9. pmid:29873683
- 6. Tobin RJ, Harrison LE, Tully MK, Lubis IND, Noviyanti R, Anstey NM, et al. Updating estimates of Plasmodium knowlesi malaria risk in response to changing land use patterns across Southeast Asia. PLoS Negl Trop Dis. 2024;18(1):e0011570. pmid:38252650
- 7. Brock PM, Fornace KM, Parmiter M, Cox J, Drakeley CJ, Ferguson HM, et al. Plasmodium knowlesi transmission: integrating quantitative approaches from epidemiology and ecology to understand malaria as a zoonosis. Parasitology. 2016;143(4):389–400. pmid:26817785
- 8. Fornace KM, Abidin TR, Alexander N, Brock P, Grigg MJ, Murphy A, et al. Association between landscape factors and spatial patterns of Plasmodium knowlesi infections in Sabah, Malaysia. Emerging Infectious Diseases. 2016;22(2):201–8.
- 9. Grigg MJ, Cox J, William T, Jelip J, Fornace KM, Brock PM, et al. Individual-level factors associated with the risk of acquiring human Plasmodium knowlesi malaria in Malaysia: a case-control study. Lancet Planet Health. 2017;1(3):e97–104. pmid:28758162
- 10. Fornace KM, Brock PM, Abidin TR, Grignard L, Herman LS, Chua TH, et al. Environmental risk factors and exposure to the zoonotic malaria parasite Plasmodium knowlesi across northern Sabah, Malaysia: a population-based cross-sectional survey. Lancet Planet Health. 2019;3(4):e179–86. pmid:31029229
- 11. Neafsey DE, Taylor AR, MacInnis BL. Advances and opportunities in malaria population genomics. Nat Rev Genet. 2021;22(8):502–17. pmid:33833443
- 12. Abdel Hamid M, Abdelraheem M, Acheampong D, Ahouidi A, Ali M, Almagro-Garcia J. Pf7: an open dataset of Plasmodium falciparum genome variation in 20,000 worldwide samples. Wellcome Open Research. 2023;8(22):22.
- 13. , Adam I, Alam MS, Alemu S, Amaratunga C, Amato R, et al. An open dataset of Plasmodium vivax genome variation in 1,895 worldwide samples. Wellcome Open Res. 2022;7:136. pmid:35651694
- 14. Benavente ED, Gomes AR, De Silva JR, Grigg M, Walker H, Barber BE. Whole genome sequencing of amplified Plasmodium knowlesi DNA from unprocessed blood reveals genetic exchange events between Malaysian Peninsular and Borneo subpopulations. Scientific Reports. 2019;9(1):.
- 15. Diez Benavente E, Florez de Sessions P, Moon RW, Holder AA, Blackman MJ, Roper C, et al. Analysis of nuclear and organellar genomes of Plasmodium knowlesi in humans reveals ancient population structure and recent recombination among host-specific subpopulations. PLoS Genet. 2017;13(9):e1007008. pmid:28922357
- 16. Pinheiro MM, Ahmed MA, Millar SB, Sanderson T, Otto TD, Lu WC, et al. Plasmodium knowlesi genome sequences from clinical isolates reveal extensive genomic dimorphism. PLoS One. 2015;10(4):e0121303. pmid:25830531
- 17. Assefa S, Lim C, Preston MD, Duffy CW, Nair MB, Adroub SA, et al. Population genomic structure and adaptation in the zoonotic malaria parasite Plasmodium knowlesi. Proc Natl Acad Sci U S A. 2015;112(42):13027–32. pmid:26438871
- 18. Turkiewicz A, Manko E, Oresegun DR, Nolder D, Spadar A, Sutherland CJ, et al. Population genetic analysis of Plasmodium knowlesi reveals differential selection and exchange events between Borneo and Peninsular sub-populations. Sci Rep. 2023;13(1):2142. pmid:36750737
- 19. Hussin N, Lim YA-L, Goh PP, William T, Jelip J, Mudin RN. Updates on malaria incidence and profile in Malaysia from 2013 to 2017. Malar J. 2020;19(1):55. pmid:32005228
- 20. Hocking SE, Divis PCS, Kadir KA, Singh B, Conway DJ. Population Genomic Structure and Recent Evolution of Plasmodium knowlesi, Peninsular Malaysia. Emerg Infect Dis. 2020;26(8):1749–58. pmid:32687018
- 21. Divis PCS, Lin LC, Rovie-Ryan JJ, Kadir KA, Anderios F, Hisam S, et al. Three Divergent Subpopulations of the Malaria Parasite Plasmodium knowlesi. Emerg Infect Dis. 2017;23(4):616–24. pmid:28322705
- 22. Divis PCS, Singh B, Anderios F, Hisam S, Matusop A, Kocken CH, et al. Admixture in Humans of Two Divergent Plasmodium knowlesi Populations Associated with Different Macaque Host Species. PLoS Pathog. 2015;11(5):e1004888. pmid:26020959
- 23.
Westaway J, Benavente E, Auburn S, Kucharski M, Aranciaga N. Plasmodium knowlesi whole genomes from Sabah, Malaysia. 2024.
- 24. Imwong M, Tanomsing N, Pukrittayakamee S, Day NPJ, White NJ, Snounou G. Spurious amplification of a Plasmodium vivax small-subunit RNA gene by use of primers currently used to detect P. knowlesi. J Clin Microbiol. 2009;47(12):4173–5. pmid:19812279
- 25. Padley D, Moody AH, Chiodini PL, Saldanha J. Use of a rapid, single-round, multiplex PCR to detect malarial parasites and identify the species present. Ann Trop Med Parasitol. 2003;97(2):131–7. pmid:12803868
- 26. Field MA, Cho V, Andrews TD, Goodnow CC. Reliably Detecting Clinically Important Variants Requires Both Combined Variant Calls and Optimized Filtering Strategies. PLoS One. 2015;10(11):e0143199. pmid:26600436
- 27. Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. Bioinformatics. 2011;27(1):3.
- 28. Benavente ED, de Sessions PF, Moon RW, Grainger M, Holder AA, Blackman MJ, et al. A reference genome and methylome for the Plasmodium knowlesi A1-H.1 line. Int J Parasitol. 2018;48(3–4):191–6. pmid:29258833
- 29.
Van der Auwera G, O’Connor B. Genomics in the Cloud: Using Docker, GATK, and WDL in Terra (1st Edition): O’Reilly Media; 2020.
- 30.
Base quality score recalibration (BQSR). 2024.
- 31. Waardenberg AJ, Field MA. consensusDE: an R package for assessing consensus of multiple RNA-seq algorithms with RUV correction. PeerJ. 2019;7e8206. pmid:31844586
- 32. Hamzeh AR, Andrews TD, Field MA. Detecting Causal Variants in Mendelian Disorders Using Whole-Genome Sequencing. Methods Mol Biol. 2021;2243:1–25. pmid:33606250
- 33. Auburn S, Campino S, Miotto O, Djimde AA, Zongo I, Manske M, et al. Characterization of within-host Plasmodium falciparum diversity using next-generation sequence data. PLoS One. 2012;7(2):e32891. pmid:22393456
- 34.
Wickham H. Elegant graphics for data analysis. 2016.
- 35. Chang CC, Chow CC, Tellier LC, Vattikuti S, Purcell SM, Lee JJ. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience. 2015;47. pmid:25722852
- 36.
Purcell S, Chang C. PLINK 2.0. n.d.
- 37. Stevens EL, Heckenberg G, Roberson EDO, Baugher JD, Downey TJ, Pevsner J. Inference of relationships in population data using identity-by-descent and identity-by-state. PLoS Genet. 2011;7(9):e1002287. pmid:21966277
- 38. Alexander DH, Novembre J, Lange K. Fast model-based estimation of ancestry in unrelated individuals. Genome Res. 2009;19(9):1655–64. pmid:19648217
- 39. Gascuel O, Steel M. Neighbor-joining revealed. Molecular Biology and Evolution. 2006;23(11):1997–2000.
- 40. Xu S, Li L, Luo X, Chen M, Tang W, Zhan L, et al. Ggtree: A serialized data object for visualization of a phylogenetic tree and annotation data. Imeta. 2022;1(4):e56. pmid:38867905
- 41. Schaffner SF, Taylor AR, Wong W, Wirth DF, Neafsey DE. hmmIBD: software to infer pairwise identity by descent between haploid genotypes. Malar J. 2018;17(1):196. pmid:29764422
- 42. Csardi G, Nepusz T. The igraph software package for complex network research. InterJournal. n.d.;Complex Systems1695.
- 43.
Hulthen A, Waha K. Indonesian landscape metrics. 2022.
- 44. Moyes CL, Shearer FM, Huang Z, Wiebe A, Gibson HS, Nijman V, et al. Predicting the geographical distributions of the macaque hosts and mosquito vectors of Plasmodium knowlesi malaria in forested and non-forested areas. Parasit Vectors. 2016;9:242. pmid:27125995
- 45. Pfeffer DA, Lucas TCD, May D, Harris J, Rozier J, Twohig KA, et al. malariaAtlas: an R interface to global malariometric data hosted by the Malaria Atlas Project. Malar J. 2018;17(1):352. pmid:30290815
- 46.
Zhou X, Lin H. Moran’s I. In: Shekhar S, Xiong H, editors. Encyclopedia of GIS. Boston, MA: Springer US; 2008. 725 p.
- 47. Cavanaugh JE, Neath AA. The Akaike information criterion: Background, derivation, properties, application, interpretation, and refinements. WIREs Computational Stats. 2019;11(3):.
- 48. Noviyanti R, Miotto O, Barry A, Marfurt J, Siegel S, Thuy-Nhien N, et al. Implementing parasite genotyping into national surveillance frameworks: feedback from control programmes and researchers in the Asia-Pacific region. Malar J. 2020;19(1):271. pmid:32718342
- 49. Benavente ED, Manko E, Phelan J, Campos M, Nolder D, Fernandez D, et al. Distinctive genetic structure and selection patterns in Plasmodium vivax from South Asia and East Africa. Nat Commun. 2021;12(1):3160. pmid:34039976
- 50. Kissinger JC, Brunk BP, Crabtree J, Fraunholz MJ, Gajria B, Milgram AJ, et al. The Plasmodium genome database. Nature. 2002;419(6906):490–2. pmid:12368860
- 51. Chen F, Mackey AJ, Stoeckert CJ Jr, Roos DS. OrthoMCL-DB: querying a comprehensive multi-species collection of ortholog groups. Nucleic Acids Res. 2006;34(Database issue):D363-8. pmid:16381887
- 52. Manske M, Miotto O, Campino S, Auburn S, Almagro-Garcia J, Maslen G, et al. Analysis of Plasmodium falciparum diversity in natural infections by deep sequencing. Nature. 2012;487(7407):375–9. pmid:22722859
- 53. van de Straat B, Sebayang B, Grigg MJ, Staunton K, Garjito TA, Vythilingam I, et al. Zoonotic malaria transmission and land use change in Southeast Asia: what is known about the vectors. Malar J. 2022;21(1):109. pmid:35361218
- 54. Gupta D, Patra A, Zhu L, Gupta A, Bozdech Z. DNA damage regulation and its role in drug-related phenotypes in the malaria parasites. Scientific Reports. 2016;6(1):23603.
- 55. Srinivasan P, Fujioka H, Jacobs-Lorena M. PbCap380, a novel oocyst capsule protein, is essential for malaria parasite survival in the mosquito. Cell Microbiol. 2008;10(6):1304–12. pmid:18248630
- 56. Kangwanrangsan N, Tachibana M, Jenwithisuk R, Tsuboi T, Riengrojpitak S, Torii M, et al. A member of the CPW-WPC protein family is expressed in and localized to the surface of developing ookinetes. Malar J. 2013;12:129. pmid:23587146
- 57. Tomley FM, Soldati DS. Mix and match modules: structure and function of microneme proteins in apicomplexan parasites. Trends Parasitol. 2001;17(2):81–8. pmid:11228014
- 58. Hayton K, Gaur D, Liu A, Takahashi J, Henschen B, Singh S, et al. Erythrocyte binding protein PfRH5 polymorphisms determine species-specific pathways of Plasmodium falciparum invasion. Cell Host Microbe. 2008;4(1):40–51. pmid:18621009
- 59. Grigg MJ, Barber BE, Marfurt J, Imwong M, William T, Bird E, et al. Dihydrofolate-Reductase Mutations in Plasmodium knowlesi Appear Unrelated to Selective Drug Pressure from Putative Human-To-Human Transmission in Sabah, Malaysia. PLoS One. 2016;11(3):e0149519. pmid:26930493
- 60. Peterson DS, Walliker D, Wellems TE. Evidence that a point mutation in dihydrofolate reductase-thymidylate synthase confers resistance to pyrimethamine in falciparum malaria. Proc Natl Acad Sci U S A. 1988;85(23):9114–8. pmid:2904149
- 61. Imwong M, Pukrittakayamee S, Looareesuwan S, Pasvol G, Poirreiz J, White NJ, et al. Association of genetic mutations in Plasmodium vivax dhfr with resistance to sulfadoxine-pyrimethamine: geographical and clinical correlates. Antimicrob Agents Chemother. 2001;45(11):3122–7. pmid:11600366
- 62. Tjitra E, Baker J, Suprianto S, Cheng Q, Anstey NM. Therapeutic efficacies of artesunate-sulfadoxine-pyrimethamine and chloroquine-sulfadoxine-pyrimethamine in vivax malaria pilot studies: relationship to Plasmodium vivax dhfr mutations. Antimicrob Agents Chemother. 2002;46(12):3947–53. pmid:12435700
- 63. Fornace KM, Topazian HM, Routledge I, Asyraf S, Jelip J, Lindblade KA, et al. No evidence of sustained nonzoonotic Plasmodium knowlesi transmission in Malaysia from modelling malaria case data. Nat Commun. 2023;14(1):2945. pmid:37263994
- 64. van Schalkwyk DA, Blasco B, Davina Nuñez R, Liew JWK, Amir A, Lau YL, et al. Plasmodium knowlesi exhibits distinct in vitro drug susceptibility profiles from those of Plasmodium falciparum. Int J Parasitol Drugs Drug Resist. 2019;993–9. pmid:30831468
- 65. Das S, Muleba M, Stevenson JC, Pringle JC, Norris DE. Beyond the entomological inoculation rate: characterizing multiple blood feeding behavior and Plasmodium falciparum multiplicity of infection in Anopheles mosquitoes in northern Zambia. Parasit Vectors. 2017;10(1):45. pmid:28122597
- 66. Wong W, Schaffner S, Thwing J, Seck M, Gomis J, Diedhiou Y, et al. Evaluating the performance of Plasmodium falciparum genetics for inferring National Malaria Control Program reported incidence in Senegal. Res Sq. 2023;Volume Number Placeholder(Issue Number Placeholder):Page Range Placeholder.
- 67. Zhu SJ, Hendry JA, Almagro-Garcia J, Pearson RD, Amato R, Miles A, et al. The origins and relatedness structure of mixed infections vary with local prevalence of P. falciparum malaria. Elife. 2019;8e40845. pmid:31298657
- 68. Kebede AM, Sutanto E, Trimarsanto H, Benavente ED, Barnes M, Pearson RD, et al. Genomic analysis of Plasmodium vivax describes patterns of connectivity and putative drivers of adaptation in Ethiopia. Scientific Reports. 2023;13(1):20788.
- 69. Sam J, Shamsusah NA, Ali AH, Hod R, Hassan MR, Agustar HK. Prevalence of simian malaria among macaques in Malaysia (2000-2021): A systematic review. PLoS Negl Trop Dis. 2022;16(7):e0010527. pmid:35849568
- 70. Moon RW, Sharaf H, Hastings CH, Ho YS, Nair MB, Rchiad Z, et al. Normocyte-binding protein required for human erythrocyte invasion by the zoonotic malaria parasite Plasmodium knowlesi. Proc Natl Acad Sci U S A. 2016;113(26):7231–6. pmid:27303038
- 71. Hawkes F, Manin B, Cooper A, Daim S, R H, Jelip J, et al. Vector compositions change across forested to deforested ecotones in emerging areas of zoonotic malaria transmission in Malaysia. Scientific Reports. 2019;9(1):13312.
- 72. Wong ML, Chua TH, Leong CS, Khaw LT, Fornace K, Wan-Sulaiman W-Y, et al. Seasonal and Spatial Dynamics of the Primary Vector of Plasmodium knowlesi within a Major Transmission Focus in Sabah, Malaysia. PLoS Negl Trop Dis. 2015;9(10):e0004135. pmid:26448052
- 73. Putaporntip C, Thongaree S, Jongwutiwes S. Differential sequence diversity at merozoite surface protein-1 locus of Plasmodium knowlesi from humans and macaques in Thailand. Infect Genet Evol. 2013;18:213–9. pmid:23727342
- 74. Saleh Huddin A, Md Yusuf N, Razak MRMA, Ogu Salim N, Hisam S. Genetic diversity of Plasmodium knowlesi among human and long-tailed macaque populations in Peninsular Malaysia: The utility of microsatellite markers. Infect Genet Evol. 2019;75:103952. pmid:31279818
- 75. Ruiz Cuenca P, Key S, Lindblade KA, Vythilingam I, Drakeley C, Fornace K. Is there evidence of sustained human-mosquito-human transmission of the zoonotic malaria Plasmodium knowlesi? A systematic literature review. Malar J. 2022;21(1):89. pmid:35300703
- 76. Fornace KM, Alexander N, Abidin TR, Brock PM, Chua TH, Vythilingam I, et al. Local human movement patterns and land use impact exposure to zoonotic malaria in Malaysian Borneo. Elife. 2019;8e47602. pmid:31638575
- 77. Stark DJ, Fornace KM, Brock PM, Abidin TR, Gilhooly L, Jalius C, et al. Long-Tailed Macaque Response to Deforestation in a Plasmodium knowlesi-Endemic Area. Ecohealth. 2019;16(4):638–46. pmid:30927165
- 78. Hans N, Singh S, Pandey AK, Reddy KS, Gaur D, Chauhan VS. Identification and characterization of a novel Plasmodium falciparum adhesin involved in erythrocyte invasion. PLoS One. 2013;8(9):e74790. pmid:24058628
- 79. Chakrabarti M, Joshi N, Kumari G, Singh P, Shoaib R, Munjal A, et al. Interaction of Plasmodium falciparum apicortin with α- and β-tubulin is critical for parasite growth and survival. Sci Rep. 2021;11(1):4688. pmid:33633135
- 80. Chakrabarti M, Garg S, Rajagopal A, Pati S, Singh S. Targeted repression of Plasmodium apicortin by host microRNA impairs malaria parasite growth and invasion. Disease Models & Mechanisms. n.d.;13(6):e04212.
- 81. Brock PM, Fornace KM, Grigg MJ, Anstey NM, William T, Cox J, et al. Predictive analysis across spatial scales links zoonotic malaria to deforestation. Proc Biol Sci. 2019;286(1894):20182351. pmid:30963872
- 82. Ang JXD, Kadir KA, Mohamad DSA, Matusop A, Divis PCS, Yaman K, et al. New vectors in northern Sarawak, Malaysian Borneo, for the zoonotic malaria parasite, Plasmodium knowlesi. Parasit Vectors. 2020;13(1):472. pmid:32933567
- 83. Byrne I, Aure W, Manin BO, Vythilingam I, Ferguson HM, Drakeley CJ, et al. Environmental and spatial risk factors for the larval habitats of Plasmodium knowlesi vectors in Sabah, Malaysian Borneo. Sci Rep. 2021;11(1):11810. pmid:34083582
- 84.
Oresegun DR, Thorpe P, Benavente ED, Campino S, Muh F, Moon RW, et al. De Novo Assembly of Plasmodium knowlesi Genomes From Clinical Samples Explains the Counterintuitive Intrachromosomal Organization of Variant SICAvar and kir Multiple Gene Family Members. Frontiers in Genetics. 2022;13.
- 85. Abdullah NR, Norahmad NA, Jelip J, Sulaiman LH, Mohd Sidek H, Ismail Z, et al. High prevalence of mutation in the Plasmodium falciparum dhfr and dhps genes in field isolates from Sabah, Northern Borneo. Malar J. 2013;12:198. pmid:23758930
- 86. Grigg MJ, William T, Menon J, Dhanaraj P, Barber BE, Wilkes CS, et al. Artesunate-mefloquine versus chloroquine for treatment of uncomplicated Plasmodium knowlesi malaria in Malaysia (ACT KNOW): an open-label, randomised controlled trial. Lancet Infect Dis. 2016;16(2):180–8. pmid:26603174
- 87. Grigg MJ, William T, Menon J, Barber BE, Wilkes CS, Rajahram GS, et al. Efficacy of Artesunate-mefloquine for Chloroquine-resistant Plasmodium vivax Malaria in Malaysia: An Open-label, Randomized, Controlled Trial. Clin Infect Dis. 2016;62(11):1403–11. pmid:27107287
- 88. Barber BE, Grigg MJ, Cooper DJ, van Schalkwyk DA, William T, Rajahram GS, et al. Clinical management of Plasmodium knowlesi malaria. Adv Parasitol. 2021;113:45–76. pmid:34620385
- 89. Moon RW, Hall J, Rangkuti F, Ho YS, Almond N, Mitchell GH, et al. Adaptation of the genetically tractable malaria pathogen Plasmodium knowlesi to continuous culture in human erythrocytes. Proc Natl Acad Sci U S A. 2013;110(2):531–6. pmid:23267069
- 90. Pain A, Böhme U, Berry AE, Mungall K, Finn RD, Jackson AP, et al. The genome of the simian and human malaria parasite Plasmodium knowlesi. Nature. 2008;455(7214):799–803. pmid:18843368