Malaria is an infectious disease caused by a microscopic parasite called Plasmodium, which is transferred between humans by mosquitos. One species of malaria parasite called Plasmodium falciparum can cause particularly severe and life-threatening forms of the disease. Currently, the most widely used treatment for P. falciparum infections is artemisinin combination therapy, a treatment that combines the drug artemisinin (or a closely related molecule) with another antimalarial drug. However, resistance to artemisinin has started to spread throughout Southeast Asia.

Artemisinin resistance is caused by mutations in a parasite gene called kelch13, and researchers have identified over 20 different mutations in P. falciparum that confer artemisinin resistance. The diversity of mutations involved, and the fact that the same mutation can arise independently in different locations, make it difficult to track the spread of resistance using conventional molecular marker approaches.

Here, Amato, Miotto et al. sequenced the entire genomes of more than 3,000 clinical samples of P. falciparum from Southeast Asia and Africa, collected as part of a global network of research groups called the MalariaGEN Plasmodium falciparum Community Project. Amato, Miotto et al. found that African parasites had independently acquired many of the same kelch13 mutations that are known to cause resistance to artemisinin in Southeast Asia. However the kelch13 mutations seen in Africa remained at low levels in the parasite population, and appeared to be under much less pressure for evolutionary selection than those found in Southeast Asia.

These findings demonstrate that the emergence and spread of resistance to antimalarial drugs does not depend solely on the mutational process, but also on other factors that influence whether the mutations will spread in the population. Understanding how this is affected by different patterns of drug treatments and other environmental conditions will be important in developing more effective strategies for combating malaria.

Artemisinin combination therapy (ACT), the frontline treatment for P. falciparum infection, has played a major part in reducing the number of deaths due to malaria over the past decade (World Health Organization, 2014b). However the increasing prevalence of artemisinin resistant P. falciparum across large parts of Southeast Asia threatens to destabilise malaria control worldwide (Dondorp et al., 2009; Hien et al., 2012; Phyo et al., 2012; Kyaw et al., 2013; Ashley et al., 2014; World Health Organization, 2014a). One of the main contemporary challenges in global health is to prevent artemisinin resistance from becoming established in Africa, where the consequences for childhood mortality could be disastrous (Dondorp and Ringwald, 2013).

Understanding the epidemiological and evolutionary driving forces behind artemisinin resistance is essential to develop effective strategies to stop it spreading. At the molecular level, artemisinin resistance is caused by mutations in a kelch protein encoded by PF3D7_1343700 on P. falciparum chromosome 13, referred to here as kelch13. More specifically, non-synonymous mutations in the kelch13 propeller and BTB-POZ domains (KPBD) result in reduced sensitivity of P. falciparum to artemisinin, as demonstrated by multiple lines of evidence including laboratory studies of artificially acquired resistance, genetic association studies of natural resistance and allelic replacement experiments (Ariey et al., 2014; Ghorbal et al., 2014; Miotto et al., 2015; Straimer et al., 2015; Takala-Harrison et al., 2015). Parasites with KBPD mutations tend to grow more slowly in the early part of the erythrocytic cycle, and have an enhanced unfolded protein response, both of which might act to protect against oxidative damage caused by artemisinin (Dogovski et al., 2015; Mok et al., 2015). It has also recently been shown that the PI3K protein is the target of artemisinin action, and that kelch13 binds to PI3K to mark it for degradation; by affecting this binding, KBPD mutations result in high levels of PI3K which counteract the effect of artemisinin (Mbengue et al., 2015).

A striking characteristic of the current distribution of artemisinin resistance is that it is caused by multiple independent KPBD mutations emerging in different locations, i.e. it does not originate from a single mutational event. More than 20 KPBD SNPs have been associated with delayed parasite clearance during artemisinin treatment and there are several documented instances of the same allele arising independently in different locations (Ashley et al., 2014; Miotto et al., 2015; Takala-Harrison et al., 2015). These are classic features of a soft selective sweep which, according to evolutionary theory, is most likely to arise in populations where the selected alleles are already present as standing genetic variations or have been repeatedly introduced by de novo mutations (Hermisson and Pennings, 2005; Pennings and Hermisson, 2006; Messer and Petrov, 2013). There is ongoing debate among evolutionary biologists about how commonly soft selective sweeps occur in nature (Jensen, 2014; Garud et al., 2015) but they have clearly played a role in previous forms of antimalarial drug resistance (Nair et al., 2007; Salgueiro et al., 2010) and the current epidemic of artemisinin resistance is the most extreme example of a soft selective sweep thus far observed in eukaryotes.

This creates a practical problem in monitoring the global spread of resistance. Artemisinin resistance can be measured directly, by following the rate of parasite clearance in patients (Flegg et al., 2011) or by testing parasite isolates in vitro (Witkowski et al., 2013), but these phenotypic assays are resource intensive and impractical for large-scale screening in resource-poor settings. Genetic approaches are therefore preferable for practical implementation of large-scale surveillance, but the soft selective sweep of artemisinin resistance produces much more heterogeneous genetic signatures than previous global waves of chloroquine and pyrimethamine resistance, where hard selective sweeps were the dominant mode of spread. Thus there is considerable uncertainty about the epidemiological significance of the growing number of non-synonymous KPBD mutations reported in Africa (Ashley et al., 2014; Hopkins Sibley, 2015; Kamau et al., 2015; Taylor et al., 2015). Previous studies in Africa have not identified variants known to cause resistance in Southeast Asia (Kamau et al., 2015; Taylor et al., 2015). At the same time, there are documented instances of African parasites that can be rapidly cleared by artemisinin treatment, although they carry KPBD mutations (Ashley et al., 2014). In the absence of comprehensive phenotypic data, it is not known which of these mutations, if any, are markers of resistance. This is a limitation of conventional molecular epidemiology, which tracks specific mutations and haplotypes and is poorly equipped to monitor soft selective sweeps where new mutations are continually arising on different haplotypic backgrounds, making it difficult to keep track of their phenotypic effects and evolutionary trajectories.

Here we explore how genomic data might help overcome these practical obstacles to monitoring the current epidemic of artemisinin resistance. This analysis includes genome sequencing data for 3,411 clinical samples of P. falciparum obtained from 43 locations in 23 countries. This large dataset was generated by the MalariaGEN Plasmodium falciparum Community Project, (www.malariagen.net/projects/parasite/pf), a collaborative study in which multiple research groups working on different scientific questions are sharing genome variation data to generate an integrated view of polymorphism in the global parasite population. Data resources arising from the present analysis, including genotype calls and sample metadata, can be obtained at www.malariagen.net/resource/16. The MalariaGEN Plasmodium falciparum Community Project also developed a user-friendly web application, with interactive tools for exploring and querying the latest version of the data: www.malariagen.net/apps/pf.

This study demonstrates the value of genomic data in characterising the current epidemic of artemisinin resistance, which is problematic for conventional molecular epidemiology since new resistance-causing mutations are continually emerging on different haplotypic backgrounds. A key problem is to define the geographical origin of KBPD mutations, and we show that this can be solved by using genomic epidemiological data to analyse ancestral relationships between samples, thereby demonstrating that the KPBD mutations observed in Africa are of local origin.

Another important question is whether KPBD mutations are under positive selection in Africa, which is difficult to determine by standard haplotype-based methods because so many independent mutations are involved. This question is further complicated by marked geographical variation in normal levels of genetic diversity, i.e. there are many more rare variants in Africa than Southeast Asia, most likely due to the larger population size and other demographic factors (Manske et al., 2012). Here we address this question by comparing kelch13 against other genes in the same samples, a process that we refer to as genomic calibration. We show that for most genes the ratio of non-synonymous to synonymous mutations is relatively constant across geographical regions, despite geographic differences in genetic diversity, and that this ratio is correlated with the level of sequence conservation across different Plasmodium species. When calibrated against other genes with the same level of cross-species conservation, allowing for geographical differences in the overall level of genetic diversity, kelch13 shows a marked excess of non-synonymous substitutions in Southeast Asia, but appears normal in Africa. Moreover, KPBD mutations causing radical amino acid changes at highly conserved positions are found at relatively high frequency in Southeast Asia but remain at very low frequency in Africa. Taken together, these findings indicate that non-synonymous KBPD mutations are undergoing strong evolutionary selection in Southeast Asia, whereas those seen in Africa have originated locally and most likely reflect normal variation.

These findings have practical implications for the prevention of artemisinin resistance in Africa, where there is evidently a deep reservoir of low frequency genetic variations that could potentially allow resistance to emerge rapidly as the levels of selective pressure increase. In most parts of Africa, the selective pressure of artemisinin is probably relatively low at present, for several reasons. Artemisinin has been widely used in Southeast Asia for over two decades, whereas its usage in Africa is more recent, and it is estimated that only 20% of infected African children currently have access to frontline treatment ACT medication (World Health Organization, 2014b). Another factor is that people living in regions of high malaria endemicity acquire partial immunity, resulting in asymptomatic infection, so that there is a large reservoir of parasites in Africa that are not exposed to antimalarial drugs because asymptomatic individuals do not seek treatment (Hastings, 2003). The situation could change dramatically as malaria control efforts are intensified, and it will be vital to monitor the effects of major interventions on the emergence of resistance, particularly in African countries that have already achieved relatively low levels of malaria transmission. A key question for the future is whether parasite populations in certain locations possess genetic features that predispose to the emergence of artemisinin resistance, as suggested by the strong association of certain fd, arps10, mdr2 and crt alleles with resistance-causing KPBD mutations in Southeast Asia (Miotto et al., 2015). Another concern is that growing resistance to ACT partner drugs, now emerging in Southeast Asia (Saunders et al., 2014) may spread to Africa, or evolve independently there, and lead to increased selective pressure for artemisinin resistance there.

The global spread of resistance to chloroquine and sulfadoxine-pyrimethamine was dominated by hard sweeps of specific haplotypes originating in Southeast Asia, although it is clear that there were also localised emergences (Mita et al., 2009). Although we still know relatively little about the functional properties of different KPBD mutations, it is clear that some are more successful than others, e.g. the C580Y allele has emerged at multiple locations in Southeast Asia and Africa, and a specific C580Y haplotype is approaching fixation in large parts of Western Cambodia (Miotto et al., 2015). The high level of sequence conservation of KPBD across Plasmodium species indicates that mutations in these domains incur fitness costs, and this is supported by the observation that although there have been multiple independent origins of resistance-causing kelch13 mutations, individual kelch13 mutations appear to have limited spread, implying that there is a substantial fitness cost in the absence of sustained drug pressure. The danger is that these fitness costs may be compensated by other genetic variants, either in kelch13 or elsewhere in the genome, and that, as a result of this continuing evolutionary process, parasites in Southeast Asia will progressively acquire higher levels of artemisinin resistance (World Health Organization, 2014a) coupled with strong biological fitness and the ability to propagate across a wide range of vector species. Under these circumstances, the current soft sweep of artemisinin resistance could give way to a pervasive hard sweep with potentially disastrous consequences.

These findings demonstrate the utility of applying genomic epidemiology to identify features of parasite demography and evolution that affect how drug resistance spreads. Future strategies to combat resistance will require better understanding of the evolutionary consequences of malaria control interventions, e.g. how the selective advantage of a resistance allele is counterbalanced by its fitness cost under different control regimes and in different geographical settings. It is now possible to approach this problem prospectively, by conducting systematic spatiotemporal sampling and genome sequencing of the parasite population as an integral part of public health interventions to prevent resistance spreading.

1. 1. MalariaGEN Plasmodium falciparum Community Project

   (2016) Sample data and catalogue of genetic variations

   Access information stated in Web page.

   http://www.malariagen.net/resource/16
2. 1. MalariaGEN Plasmodium falciparum Community Project

   (2016) Plasmodium falciparum natural genome variation

   Publicly available at European Nucleotide Archive (accession no. ERP000190).

   http://www.ebi.ac.uk/ena/data/view/ERP000190
3. 1. MalariaGEN Plasmodium falciparum Community Project

   (2016) Plasmodium falciparum Illumina sequencing R&D

   Publicly available at the European Nucleotide Archive (accession no. ERP000197).

   http://www.ebi.ac.uk/ena/data/view/ERP000197

1. 1. Adams J
   2. Kelso R
   3. Cooley L

   (2000) The kelch repeat superfamily of proteins: propellers of cell function

   Trends in Cell Biology 10:17–24.

   https://doi.org/10.1016/S0962-8924(99)01673-6

   • Google Scholar
2. 1. Ariey F
   2. Witkowski B
   3. Amaratunga C
   4. Beghain J
   5. Langlois A-C
   6. Khim N
   7. Kim S
   8. Duru V
   9. Bouchier C
   10. Ma L
   11. Lim P
   12. Leang R
   13. Duong S
   14. Sreng S
   15. Suon S
   16. Chuor CM
   17. Bout DM
   18. Ménard S
   19. Rogers WO
   20. Genton B
   21. Fandeur T
   22. Miotto O
   23. Ringwald P
   24. Le Bras J
   25. Berry A
   26. Barale J-C
   27. Fairhurst RM
   28. Benoit-Vical F
   29. Mercereau-Puijalon O
   30. Ménard D

   (2014) A molecular marker of artemisinin-resistant plasmodium falciparum malaria

   Nature 505:50–55.

   https://doi.org/10.1038/nature12876

   • Google Scholar
3. 1. Ashley EA
   2. Dhorda M
   3. Fairhurst RM
   4. Amaratunga C
   5. Lim P
   6. Suon S
   7. Sreng S
   8. Anderson JM
   9. Mao S
   10. Sam B
   11. Sopha C
   12. Chuor CM
   13. Nguon C
   14. Sovannaroth S
   15. Pukrittayakamee S
   16. Jittamala P
   17. Chotivanich K
   18. Chutasmit K
   19. Suchatsoonthorn C
   20. Runcharoen R
   21. Hien TT
   22. Thuy-Nhien NT
   23. Thanh NV
   24. Phu NH
   25. Htut Y
   26. Han KT
   27. Aye KH
   28. Mokuolu OA
   29. Olaosebikan RR
   30. Folaranmi OO
   31. Mayxay M
   32. Khanthavong M
   33. Hongvanthong B
   34. Newton PN
   35. Onyamboko MA
   36. Fanello CI
   37. Tshefu AK
   38. Mishra N
   39. Valecha N
   40. Phyo AP
   41. Nosten F
   42. Yi P
   43. Tripura R
   44. Borrmann S
   45. Bashraheil M
   46. Peshu J
   47. Faiz MA
   48. Ghose A
   49. Hossain MA
   50. Samad R
   51. Rahman MR
   52. Hasan MM
   53. Islam A
   54. Miotto O
   55. Amato R
   56. MacInnis B
   57. Stalker J
   58. Kwiatkowski DP
   59. Bozdech Z
   60. Jeeyapant A
   61. Cheah PY
   62. Sakulthaew T
   63. Chalk J
   64. Intharabut B
   65. Silamut K
   66. Lee SJ
   67. Vihokhern B
   68. Kunasol C
   69. Imwong M
   70. Tarning J
   71. Taylor WJ
   72. Yeung S
   73. Woodrow CJ
   74. Flegg JA
   75. Das D
   76. Smith J
   77. Venkatesan M
   78. Plowe CV
   79. Stepniewska K
   80. Guerin PJ
   81. Dondorp AM
   82. Day NP
   83. White NJ
   84. Tracking Resistance to Artemisinin Collaboration (TRAC)
   85. Tracking Resistance to Artemisinin Collaboration (TRAC)Tracking Resistance to Artemisinin Collaboration (TRAC)

   (2014) Spread of artemisinin resistance in plasmodium falciparum malaria

   New England Journal of Medicine 371:411–423.

   https://doi.org/10.1056/NEJMoa1314981

   • Google Scholar
4. 1. Auburn S
   2. Campino S
   3. Clark TG
   4. Djimde AA
   5. Zongo I
   6. Pinches R
   7. Manske M
   8. Mangano V
   9. Alcock D
   10. Anastasi E
   11. Maslen G
   12. MacInnis B
   13. Rockett K
   14. Modiano D
   15. Newbold CI
   16. Doumbo OK
   17. Ouédraogo JB
   18. Kwiatkowski DP
   19. Dimopoulos G

   (2011) An effective method to purify plasmodium falciparum DNA directly from clinical blood samples for whole genome high-throughput sequencing

   PLoS ONE 6:e22213.

   https://doi.org/10.1371/journal.pone.0022213

   • Google Scholar
5. 1. Bentley DR
   2. Balasubramanian S
   3. Swerdlow HP
   4. Smith GP
   5. Milton J
   6. Brown CG
   7. Hall KP
   8. Evers DJ
   9. Barnes CL
   10. Bignell HR
   11. Boutell JM
   12. Bryant J
   13. Carter RJ
   14. Keira Cheetham R
   15. Cox AJ
   16. Ellis DJ
   17. Flatbush MR
   18. Gormley NA
   19. Humphray SJ
   20. Irving LJ
   21. Karbelashvili MS
   22. Kirk SM
   23. Li H
   24. Liu X
   25. Maisinger KS
   26. Murray LJ
   27. Obradovic B
   28. Ost T
   29. Parkinson ML
   30. Pratt MR
   31. Rasolonjatovo IMJ
   32. Reed MT
   33. Rigatti R
   34. Rodighiero C
   35. Ross MT
   36. Sabot A
   37. Sankar SV
   38. Scally A
   39. Schroth GP
   40. Smith ME
   41. Smith VP
   42. Spiridou A
   43. Torrance PE
   44. Tzonev SS
   45. Vermaas EH
   46. Walter K
   47. Wu X
   48. Zhang L
   49. Alam MD
   50. Anastasi C
   51. Aniebo IC
   52. Bailey DMD
   53. Bancarz IR
   54. Banerjee S
   55. Barbour SG
   56. Baybayan PA
   57. Benoit VA
   58. Benson KF
   59. Bevis C
   60. Black PJ
   61. Boodhun A
   62. Brennan JS
   63. Bridgham JA
   64. Brown RC
   65. Brown AA
   66. Buermann DH
   67. Bundu AA
   68. Burrows JC
   69. Carter NP
   70. Castillo N
   71. Chiara E. Catenazzi M
   72. Chang S
   73. Neil Cooley R
   74. Crake NR
   75. Dada OO
   76. Diakoumakos KD
   77. Dominguez-Fernandez B
   78. Earnshaw DJ
   79. Egbujor UC
   80. Elmore DW
   81. Etchin SS
   82. Ewan MR
   83. Fedurco M
   84. Fraser LJ
   85. Fuentes Fajardo KV
   86. Scott Furey W
   87. George D
   88. Gietzen KJ
   89. Goddard CP
   90. Golda GS
   91. Granieri PA
   92. Green DE
   93. Gustafson DL
   94. Hansen NF
   95. Harnish K
   96. Haudenschild CD
   97. Heyer NI
   98. Hims MM
   99. Ho JT
   100. Horgan AM
   101. Hoschler K
   102. Hurwitz S
   103. Ivanov DV
   104. Johnson MQ
   105. James T
   106. Huw Jones TA
   107. Kang G-D
   108. Kerelska TH
   109. Kersey AD
   110. Khrebtukova I
   111. Kindwall AP
   112. Kingsbury Z
   113. Kokko-Gonzales PI
   114. Kumar A
   115. Laurent MA
   116. Lawley CT
   117. Lee SE
   118. Lee X
   119. Liao AK
   120. Loch JA
   121. Lok M
   122. Luo S
   123. Mammen RM
   124. Martin JW
   125. McCauley PG
   126. McNitt P
   127. Mehta P
   128. Moon KW
   129. Mullens JW
   130. Newington T
   131. Ning Z
   132. Ling Ng B
   133. Novo SM
   134. O’Neill MJ
   135. Osborne MA
   136. Osnowski A
   137. Ostadan O
   138. Paraschos LL
   139. Pickering L
   140. Pike AC
   141. Pike AC
   142. Chris Pinkard D
   143. Pliskin DP
   144. Podhasky J
   145. Quijano VJ
   146. Raczy C
   147. Rae VH
   148. Rawlings SR
   149. Chiva Rodriguez A
   150. Roe PM
   151. Rogers J
   152. Rogert Bacigalupo MC
   153. Romanov N
   154. Romieu A
   155. Roth RK
   156. Rourke NJ
   157. Ruediger ST
   158. Rusman E
   159. Sanches-Kuiper RM
   160. Schenker MR
   161. Seoane JM
   162. Shaw RJ
   163. Shiver MK
   164. Short SW
   165. Sizto NL
   166. Sluis JP
   167. Smith MA
   168. Ernest Sohna Sohna J
   169. Spence EJ
   170. Stevens K
   171. Sutton N
   172. Szajkowski L
   173. Tregidgo CL
   174. Turcatti G
   175. vandeVondele S
   176. Verhovsky Y
   177. Virk SM
   178. Wakelin S
   179. Walcott GC
   180. Wang J
   181. Worsley GJ
   182. Yan J
   183. Yau L
   184. Zuerlein M
   185. Rogers J
   186. Mullikin JC
   187. Hurles ME
   188. McCooke NJ
   189. West JS
   190. Oaks FL
   191. Lundberg PL
   192. Klenerman D
   193. Durbin R
   194. Smith AJ

   (2008) Accurate whole human genome sequencing using reversible terminator chemistry

   Nature 456:53–59.

   https://doi.org/10.1038/nature07517

   • Google Scholar
6. 1. Brick K
   2. Pizzi E

   (2008) A novel series of compositionally biased substitution matrices for comparing plasmodium proteins

   BMC Bioinformatics 9:236.

   https://doi.org/10.1186/1471-2105-9-236

   • Google Scholar
7. 1. Claessens A
   2. Hamilton WL
   3. Kekre M
   4. Otto TD
   5. Faizullabhoy A
   6. Rayner JC
   7. Kwiatkowski D
   8. Deitsch K

   (2014) Generation of antigenic diversity in plasmodium falciparum by structured rearrangement of var genes during mitosis

   PLoS Genetics 10:e1004812.

   https://doi.org/10.1371/journal.pgen.1004812

   • Google Scholar
8. 1. Dogovski C
   2. Xie SC
   3. Burgio G
   4. Bridgford J
   5. Mok S
   6. McCaw JM
   7. Chotivanich K
   8. Kenny S
   9. Gnädig N
   10. Straimer J
   11. Bozdech Z
   12. Fidock DA
   13. Simpson JA
   14. Dondorp AM
   15. Foote S
   16. Klonis N
   17. Tilley L
   18. Schneider DS

   (2015) Targeting the cell stress response of plasmodium falciparum to overcome artemisinin resistance

   PLOS Biology 13:e1002132.

   https://doi.org/10.1371/journal.pbio.1002132

   • Google Scholar
9. 1. Dondorp AM
   2. Nosten F
   3. Yi P
   4. Das D
   5. Phyo AP
   6. Tarning J
   7. Lwin KM
   8. Ariey F
   9. Hanpithakpong W
   10. Lee SJ
   11. Ringwald P
   12. Silamut K
   13. Imwong M
   14. Chotivanich K
   15. Lim P
   16. Herdman T
   17. An SS
   18. Yeung S
   19. Singhasivanon P
   20. Day NPJ
   21. Lindegardh N
   22. Socheat D
   23. White NJ

   (2009) Artemisinin resistance in plasmodium falciparum malaria

   New England Journal of Medicine 361:455–467.

   https://doi.org/10.1056/NEJMoa0808859

   • Google Scholar
10. 1. Dondorp AM
    2. Ringwald P

    (2013) Artemisinin resistance is a clear and present danger

    Trends in Parasitology 29:359–360.

    https://doi.org/10.1016/j.pt.2013.05.005

    • Google Scholar
11. 1. Escalante AA
    2. Ayala FJ

    (1995) Evolutionary origin of plasmodium and other apicomplexa based on rRNA genes

    Proceedings of the National Academy of Sciences of the United States of America 92:5793–5797.

    https://doi.org/10.1073/pnas.92.13.5793

    • Google Scholar
12. 1. Flegg JA
    2. Guerin PJ
    3. White NJ
    4. Stepniewska K

    (2011) Standardizing the measurement of parasite clearance in falciparum malaria: the parasite clearance estimator

    Malaria Journal 10:339.

    https://doi.org/10.1186/1475-2875-10-339

    • Google Scholar
13. 1. Gardner KB
    2. Sinha I
    3. Bustamante LY
    4. Day NPJ
    5. White NJ
    6. Woodrow CJ

    (2011) Protein-based signatures of functional evolution in plasmodium falciparum

    BMC Evolutionary Biology 11:257.

    https://doi.org/10.1186/1471-2148-11-257

    • Google Scholar
14. 1. Garud NR
    2. Messer PW
    3. Buzbas EO
    4. Petrov DA
    5. Copenhaver GP

    (2015) Recent selective sweeps in north american drosophila melanogaster show signatures of soft sweeps

    PLOS Genetics 11:e1005004.

    https://doi.org/10.1371/journal.pgen.1005004

    • Google Scholar
15. 1. Ghorbal M
    2. Gorman M
    3. Macpherson CR
    4. Martins RM
    5. Scherf A
    6. Lopez-Rubio J-J

    (2014) Genome editing in the human malaria parasite plasmodium falciparum using the CRISPR-Cas9 system

    Nature Biotechnology 32:819–821.

    https://doi.org/10.1038/nbt.2925

    • Google Scholar
16. 1. Gower JC

    (1966) Some distance properties of latent root and vector methods used in multivariate analysis

    Biometrika 53:325–338.

    https://doi.org/10.1093/biomet/53.3-4.325

    • Google Scholar
17. 1. Hastings IM

    (2003) Malaria control and the evolution of drug resistance: an intriguing link

    Trends in Parasitology 19:70–73.

    https://doi.org/10.1016/S1471-4922(02)00017-X

    • Google Scholar
18. 1. Henikoff S
    2. Henikoff JG

    (1992) Amino acid substitution matrices from protein blocks

    Proceedings of the National Academy of Sciences of the United States of America 89:10915–10919.

    https://doi.org/10.1073/pnas.89.22.10915

    • Google Scholar
19. 1. Hermisson J
    2. Pennings PS

    (2005) Soft sweeps: molecular population genetics of adaptation from standing genetic variation

    Genetics 169:2335–2352.

    https://doi.org/10.1534/genetics.104.036947

    • Google Scholar
20. 1. Hien T
    2. Thuy-Nhien N
    3. Phu N
    4. Boni MF
    5. Thanh N
    6. Nha-Ca N
    7. Thai L
    8. Thai C
    9. Van Toi P
    10. Thuan P
    11. Long L
    12. Dong L
    13. Merson L
    14. Dolecek C
    15. Stepniewska K
    16. Ringwald P
    17. White NJ
    18. Farrar J
    19. Wolbers M

    (2012) In vivo susceptibility of plasmodium falciparum to artesunate in binh phuoc province, vietnam

    Malaria Journal 11:355.

    https://doi.org/10.1186/1475-2875-11-355

    • Google Scholar
21. 1. Hopkins Sibley C

    (2015) Artemisinin resistance: the more we know, the more complicated it appears

    Journal of Infectious Diseases 211:667–669.

    https://doi.org/10.1093/infdis/jiu469

    • Google Scholar
22. 1. Jensen JD

    (2014) On the unfounded enthusiasm for soft selective sweeps

    Nature Communications 5:5281.

    https://doi.org/10.1038/ncomms6281

    • Google Scholar
23. 1. Joy DA
    2. Feng X
    3. Mu J
    4. Furuya T
    5. Chotivanich K
    6. Krettli AU
    7. Ho M
    8. Wang A
    9. White NJ
    10. Suh E
    11. Beerli P
    12. Su XZ

    (2003) Early origin and recent expansion of plasmodium falciparum

    Science 300:318–321.

    https://doi.org/10.1126/science.1081449

    • Google Scholar
24. 1. Kamau E
    2. Campino S
    3. Amenga-Etego L
    4. Drury E
    5. Ishengoma D
    6. Johnson K
    7. Mumba D
    8. Kekre M
    9. Yavo W
    10. Mead D
    11. Bouyou-Akotet M
    12. Apinjoh T
    13. Golassa L
    14. Randrianarivelojosia M
    15. Andagalu B
    16. Maiga-Ascofare O
    17. Amambua-Ngwa A
    18. Tindana P
    19. Ghansah A
    20. MacInnis B
    21. Kwiatkowski D
    22. Djimde AA

    (2015) K13-propeller polymorphisms in plasmodium falciparum parasites from sub-saharan africa

    The Journal of Infectious Diseases 211:1352–1355.

    https://doi.org/10.1093/infdis/jiu608

    • Google Scholar
25. 1. Kyaw MP
    2. Nyunt MH
    3. Chit K
    4. Aye MM
    5. Aye KH
    6. Aye MM
    7. Lindegardh N
    8. Tarning J
    9. Imwong M
    10. Jacob CG
    11. Rasmussen C
    12. Perin J
    13. Ringwald P
    14. Nyunt MM
    15. Borrmann S

    (2013) Reduced susceptibility of plasmodium falciparum to artesunate in southern myanmar

    PLoS ONE 8:e57689.

    https://doi.org/10.1371/journal.pone.0057689

    • Google Scholar
26. 1. Kyte J
    2. Doolittle RF

    (1982) A simple method for displaying the hydropathic character of a protein

    Journal of Molecular Biology 157:105–132.

    https://doi.org/10.1016/0022-2836(82)90515-0

    • Google Scholar
27. 1. Larkin MA
    2. Blackshields G
    3. Brown NP
    4. Chenna R
    5. McGettigan PA
    6. McWilliam H
    7. Valentin F
    8. Wallace IM
    9. Wilm A
    10. Lopez R
    11. Thompson JD
    12. Gibson TJ
    13. Higgins DG

    (2007) Clustal w and clustal x version 2.0

    Bioinformatics 23:2947–2948.

    https://doi.org/10.1093/bioinformatics/btm404

    • Google Scholar
28. 1. Lawson DJ
    2. Hellenthal G
    3. Myers S
    4. Falush D
    5. Copenhaver GP

    (2012) Inference of population structure using dense haplotype data

    PLoS Genetics 8:e1002453.

    https://doi.org/10.1371/journal.pgen.1002453

    • Google Scholar
29. 1. Li H
    2. Durbin R

    (2009) Fast and accurate short read alignment with burrows-wheeler transform

    Bioinformatics 25:1754–1760.

    https://doi.org/10.1093/bioinformatics/btp324

    • Google Scholar
30. 1. Manske HM
    2. Kwiatkowski DP

    (2009) SNP-o-matic

    Bioinformatics 25:2434–2435.

    https://doi.org/10.1093/bioinformatics/btp403

    • Google Scholar
31. 1. Manske M
    2. Miotto O
    3. Campino S
    4. Auburn S
    5. Almagro-Garcia J
    6. Maslen G
    7. O’Brien J
    8. Djimde A
    9. Doumbo O
    10. Zongo I
    11. Ouedraogo J-B
    12. Michon P
    13. Mueller I
    14. Siba P
    15. Nzila A
    16. Borrmann S
    17. Kiara SM
    18. Marsh K
    19. Jiang H
    20. Su X-Z
    21. Amaratunga C
    22. Fairhurst R
    23. Socheat D
    24. Nosten F
    25. Imwong M
    26. White NJ
    27. Sanders M
    28. Anastasi E
    29. Alcock D
    30. Drury E
    31. Oyola S
    32. Quail MA
    33. Turner DJ
    34. Ruano-Rubio V
    35. Jyothi D
    36. Amenga-Etego L
    37. Hubbart C
    38. Jeffreys A
    39. Rowlands K
    40. Sutherland C
    41. Roper C
    42. Mangano V
    43. Modiano D
    44. Tan JC
    45. Ferdig MT
    46. Amambua-Ngwa A
    47. Conway DJ
    48. Takala-Harrison S
    49. Plowe CV
    50. Rayner JC
    51. Rockett KA
    52. Clark TG
    53. Newbold CI
    54. Berriman M
    55. MacInnis B
    56. Kwiatkowski DP

    (2012) Analysis of plasmodium falciparum diversity in natural infections by deep sequencing

    Nature 487:375–379.

    https://doi.org/10.1038/nature11174

    • Google Scholar
32. 1. Mbengue A
    2. Bhattacharjee S
    3. Pandharkar T
    4. Liu H
    5. Estiu G
    6. Stahelin RV
    7. Rizk SS
    8. Njimoh DL
    9. Ryan Y
    10. Chotivanich K
    11. Nguon C
    12. Ghorbal M
    13. Lopez-Rubio J-J
    14. Pfrender M
    15. Emrich S
    16. Mohandas N
    17. Dondorp AM
    18. Wiest O
    19. Haldar K

    (2015) A molecular mechanism of artemisinin resistance in plasmodium falciparum malaria

    Nature 520:683–687.

    https://doi.org/10.1038/nature14412

    • Google Scholar
33. 1. Messer PW
    2. Petrov DA

    (2013) Population genomics of rapid adaptation by soft selective sweeps

    Trends in Ecology & Evolution 28:659–669.

    https://doi.org/10.1016/j.tree.2013.08.003

    • Google Scholar
34. 1. Miotto O
    2. Almagro-Garcia J
    3. Manske M
    4. MacInnis B
    5. Campino S
    6. Rockett KA
    7. Amaratunga C
    8. Lim P
    9. Suon S
    10. Sreng S
    11. Anderson JM
    12. Duong S
    13. Nguon C
    14. Chuor CM
    15. Saunders D
    16. Se Y
    17. Lon C
    18. Fukuda MM
    19. Amenga-Etego L
    20. Hodgson AVO
    21. Asoala V
    22. Imwong M
    23. Takala-Harrison S
    24. Nosten F
    25. Su Xin-zhuan
    26. Ringwald P
    27. Ariey F
    28. Dolecek C
    29. Hien TT
    30. Boni MF
    31. Thai CQ
    32. Amambua-Ngwa A
    33. Conway DJ
    34. Djimdé AA
    35. Doumbo OK
    36. Zongo I
    37. Ouedraogo J-B
    38. Alcock D
    39. Drury E
    40. Auburn S
    41. Koch O
    42. Sanders M
    43. Hubbart C
    44. Maslen G
    45. Ruano-Rubio V
    46. Jyothi D
    47. Miles A
    48. O'Brien J
    49. Gamble C
    50. Oyola SO
    51. Rayner JC
    52. Newbold CI
    53. Berriman M
    54. Spencer CCA
    55. McVean G
    56. Day NP
    57. White NJ
    58. Bethell D
    59. Dondorp AM
    60. Plowe CV
    61. Fairhurst RM
    62. Kwiatkowski DP

    (2013) Multiple populations of artemisinin-resistant plasmodium falciparum in cambodia

    Nature Genetics 45:648–655.

    https://doi.org/10.1038/ng.2624

    • Google Scholar
35. 1. Miotto O
    2. Amato R
    3. Ashley EA
    4. MacInnis B
    5. Almagro-Garcia J
    6. Amaratunga C
    7. Lim P
    8. Mead D
    9. Oyola SO
    10. Dhorda M
    11. Imwong M
    12. Woodrow C
    13. Manske M
    14. Stalker J
    15. Drury E
    16. Campino S
    17. Amenga-Etego L
    18. Thanh T-NN
    19. Tran HT
    20. Ringwald P
    21. Bethell D
    22. Nosten F
    23. Phyo AP
    24. Pukrittayakamee S
    25. Chotivanich K
    26. Chuor CM
    27. Nguon C
    28. Suon S
    29. Sreng S
    30. Newton PN
    31. Mayxay M
    32. Khanthavong M
    33. Hongvanthong B
    34. Htut Y
    35. Han KT
    36. Kyaw MP
    37. Faiz MA
    38. Fanello CI
    39. Onyamboko M
    40. Mokuolu OA
    41. Jacob CG
    42. Takala-Harrison S
    43. Plowe CV
    44. Day NP
    45. Dondorp AM
    46. Spencer CCA
    47. McVean G
    48. Fairhurst RM
    49. White NJ
    50. Kwiatkowski DP

    (2015) Genetic architecture of artemisinin-resistant plasmodium falciparum

    Nature Genetics 47:226–234.

    https://doi.org/10.1038/ng.3189

    • Google Scholar
36. 1. Mita T
    2. Tanabe K
    3. Kita K

    (2009) Spread and evolution of plasmodium falciparum drug resistance

    Parasitology International 58:201–209.

    https://doi.org/10.1016/j.parint.2009.04.004

    • Google Scholar
37. 1. Mok S
    2. Ashley EA
    3. Ferreira PE
    4. Zhu L
    5. Lin Z
    6. Yeo T
    7. Chotivanich K
    8. Imwong M
    9. Pukrittayakamee S
    10. Dhorda M
    11. Nguon C
    12. Lim P
    13. Amaratunga C
    14. Suon S
    15. Hien TT
    16. Htut Y
    17. Faiz MA
    18. Onyamboko MA
    19. Mayxay M
    20. Newton PN
    21. Tripura R
    22. Woodrow CJ
    23. Miotto O
    24. Kwiatkowski DP
    25. Nosten F
    26. Day NPJ
    27. Preiser PR
    28. White NJ
    29. Dondorp AM
    30. Fairhurst RM
    31. Bozdech Z

    (2015) Population transcriptomics of human malaria parasites reveals the mechanism of artemisinin resistance

    Science 347:431–435.

    https://doi.org/10.1126/science.1260403

    • Google Scholar
38. 1. Mu J
    2. Awadalla P
    3. Duan J
    4. McGee KM
    5. Joy DA
    6. McVean GAT
    7. Su Xin-zhuan
    8. Hey J

    (2005) Recombination hotspots and population structure in plasmodium falciparum

    PLoS Biology 3:e335.

    https://doi.org/10.1371/journal.pbio.0030335

    • Google Scholar
39. 1. Nair S
    2. Nash D
    3. Sudimack D
    4. Jaidee A
    5. Barends M
    6. Uhlemann A-C
    7. Krishna S
    8. Nosten F
    9. Anderson TJC

    (2007) Recurrent gene amplification and soft selective sweeps during evolution of multidrug resistance in malaria parasites

    Molecular Biology and Evolution 24:562–573.

    https://doi.org/10.1093/molbev/msl185

    • Google Scholar
40. 1. Nielsen R
    2. Hubisz MJ
    3. Hellmann I
    4. Torgerson D
    5. Andres AM
    6. Albrechtsen A
    7. Gutenkunst R
    8. Adams MD
    9. Cargill M
    10. Boyko A
    11. Indap A
    12. Bustamante CD
    13. Clark AG

    (2009) Darwinian and demographic forces affecting human protein coding genes

    Genome Research 19:838–849.

    https://doi.org/10.1101/gr.088336.108

    • Google Scholar
41. 1. Pennings PS
    2. Hermisson J

    (2006) Soft sweeps II--molecular population genetics of adaptation from recurrent mutation or migration

    Molecular Biology and Evolution 23:1076–1084.

    https://doi.org/10.1093/molbev/msj117

    • Google Scholar
42. 1. Phyo AP
    2. Nkhoma S
    3. Stepniewska K
    4. Ashley EA
    5. Nair S
    6. McGready R
    7. ler Moo C
    8. Al-Saai S
    9. Dondorp AM
    10. Lwin KM
    11. Singhasivanon P
    12. Day NPJ
    13. White NJ
    14. Anderson TJC
    15. Nosten F

    (2012) Emergence of artemisinin-resistant malaria on the western border of thailand: a longitudinal study

    The Lancet 379:1960–1966.

    https://doi.org/10.1016/S0140-6736(12)60484-X

    • Google Scholar
43. 1. Prugnolle F
    2. Durand P
    3. Renaud F
    4. Rousset F

    (2010) Effective size of the hierarchically structured populations of the agent of malaria: a coalescent-based model

    Heredity 104:371–377.

    https://doi.org/10.1038/hdy.2009.127

    • Google Scholar
44. 1. Salgueiro P
    2. Vicente JL
    3. Ferreira C
    4. Teófilo V
    5. Galvão A
    6. do Rosário VE
    7. Cravo P
    8. Pinto J

    (2010) Tracing the origins and signatures of selection of antifolate resistance in island populations of plasmodium falciparum

    BMC Infectious Diseases 10:163.

    https://doi.org/10.1186/1471-2334-10-163

    • Google Scholar
45. 1. Saunders DL
    2. Vanachayangkul P
    3. Lon C
    4. U.S. Army Military Malaria Research Program
    5. National Center for Parasitology, Entomology, and Malaria Control (CNM)
    6. Royal Cambodian Armed Forces

    (2014) Dihydroartemisinin–piperaquine failure in cambodia

    New England Journal of Medicine 371:484–485.

    https://doi.org/10.1056/NEJMc1403007

    • Google Scholar
46. 1. Straimer J
    2. Gnadig NF
    3. Witkowski B
    4. Amaratunga C
    5. Duru V
    6. Ramadani AP
    7. Dacheux M
    8. Khim N
    9. Zhang L
    10. Lam S
    11. Gregory PD
    12. Urnov FD
    13. Mercereau-Puijalon O
    14. Benoit-Vical F
    15. Fairhurst RM
    16. Menard D
    17. Fidock DA

    (2015) K13-propeller mutations confer artemisinin resistance in plasmodium falciparum clinical isolates

    Science 347:428–431.

    https://doi.org/10.1126/science.1260867

    • Google Scholar
47. 1. Takala-Harrison S
    2. Jacob CG
    3. Arze C
    4. Cummings MP
    5. Silva JC
    6. Dondorp AM
    7. Fukuda MM
    8. Hien TT
    9. Mayxay M
    10. Noedl H
    11. Nosten F
    12. Kyaw MP
    13. Nhien NTT
    14. Imwong M
    15. Bethell D
    16. Se Y
    17. Lon C
    18. Tyner SD
    19. Saunders DL
    20. Ariey F
    21. Mercereau-Puijalon O
    22. Menard D
    23. Newton PN
    24. Khanthavong M
    25. Hongvanthong B
    26. Starzengruber P
    27. Fuehrer H-P
    28. Swoboda P
    29. Khan WA
    30. Phyo AP
    31. Nyunt MM
    32. Nyunt MH
    33. Brown TS
    34. Adams M
    35. Pepin CS
    36. Bailey J
    37. Tan JC
    38. Ferdig MT
    39. Clark TG
    40. Miotto O
    41. MacInnis B
    42. Kwiatkowski DP
    43. White NJ
    44. Ringwald P
    45. Plowe CV

    (2015) Independent emergence of artemisinin resistance mutations among plasmodium falciparum in southeast asia

    Journal of Infectious Diseases 211:670–679.

    https://doi.org/10.1093/infdis/jiu491

    • Google Scholar
48. 1. Tanabe K
    2. Mita T
    3. Jombart T
    4. Eriksson A
    5. Horibe S
    6. Palacpac N
    7. Ranford-Cartwright L
    8. Sawai H
    9. Sakihama N
    10. Ohmae H
    11. Nakamura M
    12. Ferreira MU
    13. Escalante AA
    14. Prugnolle F
    15. Björkman A
    16. Färnert A
    17. Kaneko A
    18. Horii T
    19. Manica A
    20. Kishino H
    21. Balloux F

    (2010) Plasmodium falciparum accompanied the human expansion out of africa

    Current Biology 20:1283–1289.

    https://doi.org/10.1016/j.cub.2010.05.053

    • Google Scholar
49. 1. Taylor SM
    2. Parobek CM
    3. DeConti DK
    4. Kayentao K
    5. Coulibaly SO
    6. Greenwood BM
    7. Tagbor H
    8. Williams J
    9. Bojang K
    10. Njie F
    11. Desai M
    12. Kariuki S
    13. Gutman J
    14. Mathanga DP
    15. Martensson A
    16. Ngasala B
    17. Conrad MD
    18. Rosenthal PJ
    19. Tshefu AK
    20. Moormann AM
    21. Vulule JM
    22. Doumbo OK
    23. ter Kuile FO
    24. Meshnick SR
    25. Bailey JA
    26. Juliano JJ

    (2015) Absence of putative artemisinin resistance mutations among plasmodium falciparum in sub-saharan africa: a molecular epidemiologic study

    Journal of Infectious Diseases 211:680–688.

    https://doi.org/10.1093/infdis/jiu467

    • Google Scholar
50. 1. Venkatesan M
    2. Amaratunga C
    3. Campino S
    4. Auburn S
    5. Koch O
    6. Lim P
    7. Uk S
    8. Socheat D
    9. Kwiatkowski DP
    10. Fairhurst RM
    11. Plowe CV

    (2012) Using CF11 cellulose columns to inexpensively and effectively remove human DNA from plasmodium falciparum-infected whole blood samples

    Malaria Journal 11:41.

    https://doi.org/10.1186/1475-2875-11-41

    • Google Scholar
51. 1. Witkowski B
    2. Amaratunga C
    3. Khim N
    4. Sreng S
    5. Chim P
    6. Kim S
    7. Lim P
    8. Mao S
    9. Sopha C
    10. Sam B
    11. Anderson JM
    12. Duong S
    13. Chuor CM
    14. Taylor WRJ
    15. Suon S
    16. Mercereau-Puijalon O
    17. Fairhurst RM
    18. Menard D

    (2013) Novel phenotypic assays for the detection of artemisinin-resistant plasmodium falciparum malaria in cambodia: in-vitro and ex-vivo drug-response studies

    The Lancet Infectious Diseases 13:1043–1049.

    https://doi.org/10.1016/S1473-3099(13)70252-4

    • Google Scholar
52. Book

    1. World Health Organization

    (2014a)

    Update on Artemisinin Resistance - September 2014

    Geneva: World Health Organization.

    • Google Scholar
53. Book

    1. World Health Organization

    (2014b)

    World Malaria Report 2014

    Geneva: World Health Organization.

    • Google Scholar

Thank you for submitting your work entitled "Genomic epidemiology of the current wave of artemisinin resistant malaria" for peer review at eLife. Your submission has been favorably evaluated by Prabhat Jha (Senior Editor) and three reviewers, one of whom is a member of our Board of Reviewing Editors.

The reviewers have discussed the reviews with one another and the Reviewing editor has drafted this decision to help you prepare a revised submission

Summary:

This study presents a large number of P. falciparum genomes that are a valuable resource for future molecular epidemiology studies. By analyzing these data, the authors demonstrate that:

1) Artimisinin resistance mutations arose many times independently in South East Asia and are associated with drastic amino acid changes in different domains of kelch13;

2) Resistance mutations in Africa arose locally;

3) Resistance mutations in kelch13 likely come at a substantial cost, limiting their spread.

In addition, they provide a population genomic characterization of the P. falciparum diversity.

Essential revisions:

1) The study is currently focused exclusively on kelch13 and lacks comparison to other known resistance loci. While kelch13 is compared to the genome wide distribution of diversity, dN/dS, etc., it should be put into context of other resistance loci. It would be straightforward to repeat some of the analysis done for kelch13 for other loci (e.g. highlight those genes in Figure 5, add trees of haplotypes surrounding these loci in analogy to Figure 2).

2) Data availability: the malariaGEN project hosts a comprehensive website for genomic data. However, in the interest of reproducibility and reusability additional files should be provided. We would like to see metadata for all strains included in this study. Ideally, this metadata include sampling date and location, phenotype information where available, the short read archive identifier. Similarly, the source data for figures, such as tables with dN/dS values for genes in case of Figure 5, should be provided. For follow-up analysis a flat file with the genotypes of all strains at the 935,601 high confidence SNPs and their allele frequencies in the different populations would be useful. Such files should be uploaded on data Dryad, or alternatively, a clear description on how to obtain such data from malariaGEN should be given.

3) Figure 1: a genome wide tree is a bad way to summarize diversity in a sexual population. A projection on the first two principal components could be more useful. Location, isolates with kelch13 mutations, phenotypes could be highlighted by color, symbol, size etc. In addition, genomic differentiation could be summarized by F_ST or the density of private SNPs, possibly in a sliding window along the genome. kelch13 and other resistance loci should be outliers here.

4) Population genomic characterization: Figure 3 is uninformative. Please show the complete minor allele frequency spectra, possibly on a log-log scale to improve readability at the low frequency end. Panel c is redundant with 5a. In order to interpret Figure 2, estimates of the extent of linkage disequilibrium would be welcome (why not show LD decay surrounding kelch13 and other resistance loci in SEA and Africa and compare to genome wide average).

5) Hydrophobicity and radical substitutions: Given that the function of kelch13 is unknown, the emphasis on hydrophobicity as a score for mutations is not justified. In fact, mutations in SEA are found at hydrophilic and hydrophobic sites alike. The stronger signal seems to come from conservation (but note the outlier Y493H). We suggest focusing on conservation and drop the hydrophobicity discussion unless you provide convincing evidence for a causal role of hydrophobicity changes (otherwise, it could stay as a somewhat speculative supplementary figure). The description of the calculation of the conservation score needs more detail. Why did you use a 9 amino acid window for smoothing? How did you average over multiple pairwise comparisons? While substitution matrices quantify broadly the exchangeability of amino acids, they are often a rather poor guide for site specific mutation effects. Is there a way to assess site specific conservation in a broader alignment than the one used for Table 5?

6) Resistance mutations in Africa: A more detailed analysis is required here. Please point out African strains that carry kelch13 mutations on the tree in Figure 2, show additional trees for different haplotype length, and maybe compare African haplotypes with resistance mutations explicitly to the closest SEA haplotype and the closest African haplotype lacking kelch13 mutations. In Figure 2b, do African isolates that cluster with the SEA ones have special properties (e.g. C580Y mutations)? In order to render the discussion of the resistance mutations observed in Africa more concrete, a supplementary file should be provided that contains their country of origin, any drug phenotype data, whether any were culture adapted, whether these were PCR resequenced to confirm the mutations, whether there were any clinical or parasitological data to indicate that these were associated with delayed clearance rates.

7) How representative is Figure 2b showing imperfect separation between Asia and Africa? How does this depend on the size of the window used for tree building? The two African samples that fall in between the SEA and African clusters should be discussed in greater detail and additional analyses are needed to clarify whether they are admixed.

8) The association of fd, aprs10, mdr2, etc. should be explicitly discussed as correlation rather than causation as no evidence is presented for the latter. Given resource limitation, the case for including these loci in routine surveillance is weak at present. The statement in the Discussion should be toned down.

We thank the referees for their helpful comments and interest in these data. Based on their suggestions we have made significant changes to all of the main figures. A key question was the geographical origin of kelch13 mutations found in Africa, and we have gone back to the drawing board and reanalysed this using an entirely new method: although the overall conclusions remain the same, our new approach is less dependent on assumptions and which gives results that are clearer and easier to present. The revised manuscript also has greater clarity about the extensive data resources that underlie this analysis, including open access web application that provides user-friendly tools to explore the properties and regional allele frequencies of all of the >900k SNPs used in this analysis, and to learn about the 31 partner studies and 43 sampling locations that contributed to this consortial project.

Summary:

This study presents a large number of P. falciparum genomes that are a valuable resource for future molecular epidemiology studies. By analyzing these data, the authors demonstrate that: 1) Artimisinin resistance mutations arose many times independently in South East Asia and are associated with drastic amino acid changes in different domains of kelch13; 2) Resistance mutations in Africa arose locally;

3) Resistance mutations in kelch13 likely come at a substantial cost, limiting their spread.

In addition, they provide a population genomic characterization of the P. falciparum diversity.

The three bullet points are a good summary of our main conclusions. However it is a bit of an overstatement to add that we “provide a population genomic characterization of P. falciparum diversity”. Although this paper does contain a lot of population genomic information, it is to provide context for our analysis of the kelch13 locus. For a more comprehensive set of population genomic analyses on these samples, readers are referred to other outputs of the P. falciparum Community Project and its companion Pf3k Project, as outlined below.

Essential revisions:

1) The study is currently focused exclusively on kelch13 and lacks comparison to other known resistance loci. While kelch13 is compared to the genome wide distribution of diversity, dN/dS, etc., it should be put into context of other resistance loci. It would be straightforward to repeat some of the analysis done for kelch13 for other loci (e.g. highlight those genes in Figure 5, add trees of haplotypes surrounding these loci in analogy to Figure 2).

We intentionally focus on kelch13 because of its scientific interest and practical importance. It would be fascinating to perform detailed analyses on other resistance loci, and then to compare and contrast the findings, but that is beyond the scope of the present paper. Different resistance loci have evolved over different timescales, with different demographic histories and selective processes, and each requires a focused analysis to understand what is going on. For example, detailed characterisation of the pfcrt locus alone would be a major undertaking, given the remarkably complex patterns of diversity that have emerged in different parts of the world following >60 years of selective pressure.

With the caveat that such comparisons can be misleading if taken out of context, in the revised version we show N/S ratios for other major resistance loci: pfcrt, pfmdr1, pfdhfr and pfdhps. It can be seen that both pfcrt and kelch13 have high N/S ratios in Southeast Asia, but readers need to be aware that very different processes are driving this signal for the two loci – the former is associated with a hard sweep and extreme loss of diversity, whereas the latter is associated with a multitude of resistant haplotypes. Because of these caveats, we propose that this new figure is supplementary to Figure 5, retaining the main focus of the paper on the kelch13 evolutionary pattern.

2) Data availability: the malariaGEN project hosts a comprehensive website for genomic data. However, in the interest of reproducibility and reusability additional files should be provided. We would like to see metadata for all strains included in this study. Ideally, this metadata include sampling date and location, phenotype information where available, the short read archive identifier. Similarly, the source data for figures, such as tables with dN/dS values for genes in case of Figure 5, should be provided. For follow-up analysis a flat file with the genotypes of all strains at the 935,601 high confidence SNPs and their allele frequencies in the different populations would be useful. Such files should be uploaded on data Dryad, or alternatively, a clear description on how to obtain such data from malariaGEN should be given.

We now provide additional supporting data, as summarised below, and the revised manuscript describes more clearly how to access these and other data that we have already released.

i. We have created a resource page at http://www.malariagen.net/resource/16 to accompany this paper, which provides links to the data release plus other information about the project. This takes the reader to a permanent URL listing all the online resources relating to this study. The revised manuscript gives links to these web resources in two places: in the introduction and in a new section entitled 'Data access and URLs'. In addition we provide links to specific resources where appropriate in the Results and Methods sections.

ii. We have released details of all high confidence SNPs analysed in this study, together with their regional allele frequencies, through a user-friendly web application with interactive tools for exploring and querying the data. This is now online at www.malariagen.net/apps/pf

iii. We have released genotype calls on individual samples (i.e. 935,601 SNPs genotyped in 3,394 samples). These are made available via the Wellcome Trust Sanger Institute public ftp site and can be accessed from http://www.malariagen.net/content/p-falciparum-community-project-jan-2016-data-release. We are unable to openly release the sequence data from Indonesia due to national export restrictions. This involves only 17 samples, i.e. less than 1% of the total.

iv. We provide an Excel spreadsheet containing the source data for Figures 3 and 5. For each P. falciparum gene analyzed, this table lists: the ortholog gene in P. chabaudi, the conservation score, the count of non-synonymous and synonymous mutation in Africa and Southeast Asia, the log fold change and the associated p-value. We propose that this is released as a supplementary file to the paper.

v. The sequence read data have been deposited in the European Nucleotide Archive (ENA). At the time of publication we will release a sample metadata file on www.malariagen.net/resource/16. Each sample will be identified by its ENA accession number, country of origin, year of collection, and by partner study that contributed the sample.

vi. Many of these samples are also part of the Pf3k Project whose goal is to undertake a comprehensive analysis of genome variation in 3,000 parasite samples representing the major malaria endemic regions of the world, and to make all of these data openly available. The Pf3k web application provides open access variant call data for 2,375 (68%) of the 3,488 samples included in the present study (www.malariagen.net/apps/pf3k) and these data will be continually improved as new analytical methods become available.

To put the above data resources in context, the present manuscript is one output of a larger Community Project with a dual purpose: (1) to provide malaria research groups with access to genome sequencing on their samples and to enable them to incorporate these data into their own analyses; (2) to combine data across multiple partner studies to perform global analyses such as the current manuscript (http://www.malariagen.net/projects/parasite/pf). One of the major challenges is to develop an equitable data access policy that protects the legitimate interests of individual partner studies while encouraging global analyses and openly accessible data. The data resources described above represent our best efforts to balance these different considerations while continually looking for ways to increase the quality and quantity of data that is made openly available to the scientific community.

3) Figure 1: a genome wide tree is a bad way to summarize diversity in a sexual population. A projection on the first two principal components could be more useful. Location, isolates with kelch13 mutations, phenotypes could be highlighted by color, symbol, size etc. In addition, genomic differentiation could be summarized by F_ST or the density of private SNPs, possibly in a sliding window along the genome. kelch13 and other resistance loci should be outliers here.

Unfortunately it is impossible to summarize the diversity of this population in a single graphic. Both neighbor-joining trees and PCA plots have limitations, and the most pragmatic approach is to use a variety of methods. In Figure 1—figure supplement 1 it can be seen that PC1 separates Africa and South America from the rest of the world, whereas PC2 is driven by recent founder events in Southeast Asia (see Miotto et al. 2013 for a fuller discussion). These founder effects can also be seen on the NJ tree but are much less pronounced, and we therefore felt that the NJ tree was more suitable for Figure 1, as it provides a more intuitive visualization for the average reader. However we do not feel strongly about this and are happy to include both NJT and PCA as main display items.

As suggested we have also included a sliding-window F_ST analysis which shows that the differentiation between Africa and SE Asia is fairly evenly distributed across the genome. At the pfcrt locus, for example, F_ST between Africa and SE Asia is relatively low, possibly reflecting the spread of chloroquine resistance haplotypes from SE Asia to Africa. Around the kelch13 locus there is a much stronger signal but closer inspection reveals that this is not due to kelch13 mutant haplotypes (which are diverse and present in only ~30% of SE Asian samples) but to nearby alleles that are close to fixation in SE Asia and presumably antedate the kelch13 soft sweep (see, for example, www.malariagen.net/apps/pf). Thus the results of sliding-window F_ST analysis are not as easy to interpret as the reviewer implies, and we would suggest not including this figure.

4) Population genomic characterization: Figure 3 is uninformative. Please show the complete minor allele frequency spectra, possibly on a log-log scale to improve readability at the low frequency end.

We have replaced Figure 3a with a more detailed plot of the minor allele frequency spectra. We use a log-linear scale to highlight the large number of rare alleles in the African population.

Panel c is redundant with 5a.

These are the same data but presented somewhat differently in different contexts to assist the narrative. We are happy to make 3c a supplementary figure if the editor wishes.

In order to interpret Figure 2, estimates of the extent of linkage disequilibrium would be welcome (why not show LD decay surrounding kelch13 and other resistance loci in SEA and Africa and compare to genome wide average).

Thanks for this perceptive comment. Genome-wide levels of LD are known to differ between Africa and Southeast Asia (Manske et al.2012), and these differences are accentuated by founder effects associated with artemisinin resistance (Miotto et al. 2013). As a result of this comment and other considerations, we decided to replace Figure 2 with a new method of analysis that does not assume a specific haplotype length and should therefore be more robust to variations in LD. This is described in more detail in comment 6.

5) Hydrophobicity and radical substitutions: Given that the function of kelch13 is unknown, the emphasis on hydrophobicity as a score for mutations is not justified. In fact, mutations in SEA are found at hydrophilic and hydrophobic sites alike. The stronger signal seems to come from conservation (but note the outlier Y493H). We suggest focusing on conservation and drop the hydrophobicity discussion unless you provide convincing evidence for a causal role of hydrophobicity changes (otherwise, it could stay as a somewhat speculative supplementary figure).

We have revised the main text to provide a clearer rationale of the potential importance of hydrophobicity. Mbengue et al. (Nature 2015) have shown that kelch13 binds to PI3K (a target of artemisinin), marking it for degradation. Kelch binding sites in general are determined by the arrangement of hydrophobic β-strands in the propeller domain. It is therefore interesting that radical mutations at hydrophobic sites are relatively common in Southeast Asia compared to Africa, or compared to variations seen between different Plasmodium species.

However we are not arguing that hydrophilic changes are unimportant, and to avoid undue emphasis on hydrophobicity, we have moved what was formerly Figure 6 to become a supplementary figure. In the new Figure 6 (formerly Figure 7) we have created two panels to show more clearly the different structural distribution of mutations in SEA and Africa. We have also removed 7b which highlighted the periodic pattern of hydrophobicity.

The description of the calculation of the conservation score needs more detail. Why did you use a 9 amino acid window for smoothing? How did you average over multiple pairwise comparisons?

We have applied the method described by Gardner et al. (BMC Evolutionary Biology 2011), which includes a 9-amino acid smoothing window. This did not involve averaging over multiple comparisons: as stated in Results: "Each gene was assigned a conservation score determined from a sequence alignment of the P. falciparum gene with its P. chabaudi homologue.” The problem with using multiple species in the genome-wide analysis is that it greatly reduces the number of genes that can be analysed due to lack of complete reference sequences for most species. In the revised manuscript we clarify this point: “P. chabaudi was chosen since it was the member of the group most differentiated from P. falciparum (rodent plasmodia) with the most complete reference sequence." However for kelch13 there are available data from multiple species, as shown in revised Figure 4b.

While substitution matrices quantify broadly the exchangeability of amino acids, they are often a rather poor guide for site specific mutation effects. Is there a way to assess site specific conservation in a broader alignment than the one used for Table 5?

Table 5 summarizes the conservation data that we can garner from alignments that capture ~50 Mya of parasite evolution. A broader alignment would require analysis of non-Plasmodium species with very low levels of sequence identity, which we think is beyond the scope of the present work.

6) Resistance mutations in Africa: A more detailed analysis is required here. Please point out African strains that carry kelch13 mutations on the tree in Figure 2, show additional trees for different haplotype length, and maybe compare African haplotypes with resistance mutations explicitly to the closest SEA haplotype and the closest African haplotype lacking kelch13 mutations. In Figure 2b, do African isolates that cluster with the SEA ones have special properties (e.g. C580Y mutations)? In order to render the discussion of the resistance mutations observed in Africa more concrete, a supplementary file should be provided that contains their country of origin, any drug phenotype data, whether any were culture adapted, whether these were PCR resequenced to confirm the mutations, whether there were any clinical or parasitological data to indicate that these were associated with delayed clearance rates. 7) How representative is Figure 2b showing imperfect separation between Asia and Africa? How does this depend on the size of the window used for tree building? The two African samples that fall in between the SEA and African clusters should be discussed in greater detail and additional analyses are needed to clarify whether they are admixed.

These are insightful comments, which along with other considerations have caused us to rethink this part of our analysis. Any flanking haplotype method that relies on a specific window size (such as the trees we showed in the previous version of Figure 2) is problematic because too-large a window may mask signals of recombination close to the gene of interest, whereas too-small a window will result in lower genetic distances making the tree topology unreliable. A further problem in choosing window size (as highlighted in Comment 4) is that LD decay is difficult to estimate in SEA.

We now use an entirely different approach to determine the most likely geographic origin of kelch13 mutations, and Figure 2 has been revised accordingly. We have conducted a new analysis using chromosome painting, which provides a probabilistic framework to identify shared haplotypes and analyse haplotype origin. As shown in the revised version of Figure 2, based on analysis of both kelch13 flanking regions, there are two African samples that could plausibly have common origin with Asian samples, while another four are similar to Asian samples in one of the flanking regions.

The revised manuscript has a more detailed discussion of these findings. The two African samples with Asian features in the region of kelch13 represent only a small minority of the African samples observed to have kelch13 mutations, and they were collected in two different countries. They are clearly separated from other African samples when analysed at the level of the whole genome (Figure 1) and to complicate matters they are also multiclonal infections. Therefore they could possibly be mixtures of African and Asian parasites, either real or due to laboratory contamination, and the present data do not allow us to confidently rule out either of these possibilities. Taken together, these findings offer no clear evidence that kelch13 mutations have spread from Asia to Africa, but this is clearly something that will need to be kept under close surveillance in the future.

In comment 2 we clarify the available metadata which includes country of origin and year of sampling. The dataset assembled for this consortial analysis described here does not include any clinical, parasitological or phenotypic data. The samples were contributed by multiple partner studies pursuing specific research questions independently of the consortial analysis, as outlined in the accompanying web application (www.malariagen.net/apps/pf/). The vast majority of samples were leukocyte-depleted venous blood, sequenced without culture adaptation. We used stringent quality filters to minimize the likelihood of false positive mutations, an approach which is at least as accurate as PCR resequencing as we have reported elsewhere (Manske et al.2012). Note that many African kelch13 mutations are found in multiclonal infections, which can be problematic for conventional PCR resequencing.

8) The association of fd, aprs10, mdr2, etc., should be explicitly discussed as correlation rather than causation as no evidence is presented for the latter. Given resource limitation, the case for including these loci in routine surveillance is weak at present. The statement in the Discussion should be toned down.

We have modified the Discussion as follows: “A key question for the future is whether parasite populations in certain locations possess genetic features that predispose to the emergence of artemisinin resistance, as suggested by the strong association of certain fd, arps10, mdr2 and crt alleles with resistance-causing KPBD mutations in Southeast Asia.”

Author details

1. MalariaGEN Plasmodium falciparum Community Project

   Contribution

   MGENPfCP, RA, OM, DPK, Conception and design, Analysis and interpretation of data, Drafting or revising the article, Approved the submitted manuscript; DMe, ED, SC, MK, MS, Developed and implemented sample processing and sequencing library preparation methods, Drafting or revising the article, Approved the submitted manuscript; AAN, CA, LAE, VA, TA, EA, SA, VB, AB, MFB, TB, PCB, DJC, AC, NPD, AD, CDo, AMD, CDr, PD, DFE, TGE, RMF, MAF, CIF, TTH, AH, MI, DI, PL, CL, JMa, KM, MMay, PM, VM, OAM, JMo, IM, MPK, PNN, FN, RN, AN, HO, AO, MO, J-BO, APPP, CP, RNP, SP, MR, PR, LR, DS, AS, PS, STH, T-NNT, VT, FV, JW, NJW, HY, GAA, SB, OB, KC, Carried out clinical studies and/or laboratory work to obtain parasite samples, Drafting or revising the article, Approved the submitted manuscript; CM, DJ, MMan, JS, Development and management of data production pipelines, Drafting or revising the article, Approved the submitted manuscript; CJW, JAG, IS, AM, RDP, PV, BJ, Developed analysis, software tools and methods, Drafting or revising the article, Approved the submitted manuscript; VJC, RG, DMu, CH, GM, KAR, BM, JCR, Contributed to study design and management, Drafting or revising the article, Approved the submitted manuscript.

   For correspondence

   1. dominic@sanger.ac.uk
   2. ra4@sanger.ac.uk
   3. olivo@tropmedres.ac

   Competing interests

   GM: Reviewing editor, eLife. The other authors declare that no competing interests exist.

   1. Roberto Amato, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   2. Olivo Miotto, Mahidol-Oxford Tropical Medicine Research Unit, Mahidol University, Bangkok, Thailand
   3. Charles J Woodrow, Mahidol-Oxford Tropical Medicine Research Unit, Mahidol University, Bangkok, Thailand
   4. Jacob Almagro-Garcia, Centre for Genomics and Global Health, Wellcome Trust Centre for Human Genetics, Oxford, United Kingdom
   5. Ipsita Sinha, Mahidol-Oxford Tropical Medicine Research Unit, Mahidol University, Bangkok, Thailand
   6. Susana Campino, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   7. Daniel Mead, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   8. Eleanor Drury, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   9. Mihir Kekre, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   10. Mandy Sanders, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   11. Alfred Amambua-Ngwa, Medical Research Council Unit, Banjul, The Gambia
   12. Chanaki Amaratunga, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rockville, United States
   13. Lucas Amenga-Etego, Navrongo Health Research Centre, Navrongo, Ghana
   14. Voahangy Andrianaranjaka, Malaria Research Unit, Institut Pasteur de Madagascar, Antananarivo, Madagascar
   15. Tobias Apinjoh, University of Buea, Buea, Cameroon
   16. Elizabeth Ashley, Mahidol-Oxford Tropical Medicine Research Unit, Mahidol University, Bangkok, Thailand
   17. Sarah Auburn, Global and Tropical Health Division, Menzies School of Health Research and Charles Darwin University, Darwin, Australia
   18. Gordon A Awandare, West African Centre for Cell Biology of Infectious Pathogens, College of Basic and Applied Sciences, University of Ghana, Legon, Ghana
   19. Vito Baraka, National Institute for Medical Research, Tanga - Centre, Tanga, Tanzania
   20. Alyssa Barry, Division of Population Health and Immunity, Walter and Eliza Hall Institute of Medical Research, Parkville, Australia
   21. Maciej F Boni, University of Oxford, Ho Chi Minh City, Vietnam
   22. Steffen Borrmann, Kenya Medical Research Institute/Wellcome Trust Collaborative Research Program, Kilifi, Kenya
   23. Teun Bousema, London School of Hygiene and Tropical Medicine, London, United Kingdom
   24. Oralee Branch, New York University School of Medicine, New York, United States
   25. Peter C Bull, Kenya Medical Research Institute/Wellcome Trust Collaborative Research Program, Kilifi, Kenya
   26. Kesinee Chotivanich, Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand
   27. David J Conway, Medical Research Council Unit, Banjul, The Gambia
   28. Alister Craig, Liverpool School of Tropical Medicine, Liverpool, United Kingdom
   29. Nicholas P Day, Mahidol-Oxford Tropical Medicine Research Unit, Mahidol University, Bangkok, Thailand
   30. Abdoulaye Djimdé, Malaria Research and Training Centre, Faculty of Medicine, University of Bamako, Bamako, Mali
   31. Christiane Dolecek, Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam
   32. Arjen M Dondorp, Mahidol-Oxford Tropical Medicine Research Unit, Mahidol University, Bangkok, Thailand
   33. Chris Drakeley, London School of Hygiene and Tropical Medicine, London, United Kingdom
   34. Patrick Duffy, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rockville, United States
   35. Diego F Echeverry, Department of Entomology, Purdue University, West Lafayette, United States
   36. Thomas G Egwang, Med Biotech Laboratories, Kampala, Uganda
   37. Rick M Fairhurst, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rockville, United States
   38. Md. Abul Faiz, Malaria Research Group and Dev Care Foundation, Dhaka, Bangladesh
   39. Caterina I Fanello, Mahidol-Oxford Tropical Medicine Research Unit, Mahidol University, Bangkok, Thailand
   40. Tran Tinh Hien, Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam
   41. Abraham Hodgson, Navrongo Health Research Centre, Navrongo, Ghana
   42. Mallika Imwong, Mahidol-Oxford Tropical Medicine Research Unit, Mahidol University, Bangkok, Thailand
   43. Deus Ishengoma, National Institute for Medical Research, Tanga - Centre, Tanga, Tanzania
   44. Pharath Lim, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rockville, United States
   45. Chanthap Lon, Department of Immunology and Medicine, US Army Medical Component, Armed Forces Research Institute of Medical Sciences, Bangkok, Thailand
   46. Jutta Marfurt, Global and Tropical Health Division, Menzies School of Health Research and Charles Darwin University, Darwin, Australia
   47. Kevin Marsh, Kenya Medical Research Institute/Wellcome Trust Collaborative Research Program, Kilifi, Kenya
   48. Mayfong Mayxay, Lao-Oxford-Mahosot Hospital-Wellcome Trust Research Unit, Vientiane, Lao PDR
   49. Pascal Michon, Faculty of Medicine and Health Sciences, Divine Word University, Madang, Papua New Guinea
   50. Victor Mobegi, Medical Research Council Unit, Banjul, The Gambia
   51. Olugbenga A Mokuolu, Department of Paediatrics and Child Health, University of Ilorin, Ilorin, Nigeria
   52. Jacqui Montgomery, Department of Biology and Entomology, Pennsylvania State University, University Park, United States
   53. Ivo Mueller, Department of Medical Biology, University of Melbourne, Carlton, Australia
   54. Myat Phone Kyaw, Department of Medical Research Lower Myanmar, Ministry of Health, Yangon, Myanmar
   55. Paul N Newton, Lao-Oxford-Mahosot Hospital-Wellcome Trust Research Unit, Vientiane, Lao PDR
   56. Francois Nosten, Mahidol-Oxford Tropical Medicine Research Unit, Mahidol University, Mae Sot, Thailand
   57. Rintis Noviyanti, Eijkman Institute for Molecular Biology, Jakarta, Indonesia
   58. Alexis Nzila, Department of Biology, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia
   59. Harold Ocholla, Malawi Liverpool Wellcome Trust Clinical Research Programme, Malawi
   60. Abraham Oduro, Navrongo Health Research Centre, Navrongo, Ghana
   61. Marie Onyamboko, University of Kinshasa, Kinshasa, Democratic Republic of Congo
   62. Jean-Bosco Ouedraogo, Institut de Recherche en Sciences de la Sante, Direction Régionale de l'Ouést, Bobo-Dioulasso, Burkina Faso
   63. Aung Pyae P Phyo, Shoklo Malaria Research Unit, Bangkok, Thailand
   64. Christopher Plowe, Institute for Global Health, University of Maryland School of Medicine, Baltimore, United States
   65. Ric N Price, Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, Oxford, United Kingdom
   66. Sasithon Pukrittayakamee, Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand
   67. Milijaona Randrianarivelojosia, Malaria Research Unit, Institut Pasteur de Madagascar, Antananarivo, Madagascar
   68. Pascal Ringwald, Global Malaria Programme, World Health Organization, Geneva, Switzerland
   69. Lastenia Ruiz, Universidad Nacional de la Amazonia Peruana, Iquitos, Peru
   70. David Saunders, Department of Immunology and Medicine, US Army Medical Component, Armed Forces Research Institute of Medical Sciences, Bangkok, Thailand
   71. Alex Shayo, Department of Biotechnology and Bioinformatics, The University of Dodoma, Dodoma, Tanzania
   72. Peter Siba, Papua New Guinea Institute of Medical Research, Madang, Papua New Guinea
   73. Shannon Takala-Harrison, Institute for Global Health, University of Maryland School of Medicine, Baltimore, United States
   74. Thuy-Nhien N Thanh, Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, Oxford, United Kingdom
   75. Vandana Thathy, Kenya Medical Research Institute/Wellcome Trust Collaborative Research Program, Kilifi, Kenya
   76. Federica Verra, London School of Hygiene and Tropical Medicine, London, United Kingdom
   77. Jason Wendler, Centre for Genomics and Global Health, Wellcome Trust Centre for Human Genetics, Oxford, United Kingdom
   78. Nicholas J White, Mahidol-Oxford Tropical Medicine Research Unit, Mahidol University, Bangkok, Thailand
   79. Htut Ye, Department of Medical Research Lower Myanmar, Ministry of Health, Yangon, Myanmar
   80. Victoria J Cornelius, Centre for Genomics and Global Health, Wellcome Trust Centre for Human Genetics, Oxford, United Kingdom
   81. Rachel Giacomantonio, Centre for Genomics and Global Health, Wellcome Trust Centre for Human Genetics, Oxford, United Kingdom
   82. Dawn Muddyman, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   83. Christa Henrichs, Centre for Genomics and Global Health, Wellcome Trust Centre for Human Genetics, Oxford, United Kingdom
   84. Cinzia Malangone, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   85. Dushyanth Jyothi, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   86. Richard D Pearson, Centre for Genomics and Global Health, Wellcome Trust Centre for Human Genetics, Oxford, United Kingdom
   87. Julian C Rayner, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   88. Gilean McVean, Centre for Genomics and Global Health, Wellcome Trust Centre for Human Genetics, Oxford, United Kingdom
   89. Kirk A Rockett, Centre for Genomics and Global Health, Wellcome Trust Centre for Human Genetics, Oxford, United Kingdom
   90. Alistair Miles, Centre for Genomics and Global Health, Wellcome Trust Centre for Human Genetics, Oxford, United Kingdom
   91. Paul Vauterin, Centre for Genomics and Global Health, Wellcome Trust Centre for Human Genetics, Oxford, United Kingdom
   92. Ben Jeffery, Centre for Genomics and Global Health, Wellcome Trust Centre for Human Genetics, Oxford, United Kingdom
   93. Magnus Manske, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   94. Jim Stalker, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   95. Bronwyn MacInnis, Wellcome Trust Sanger Institute, Cambridge, United Kingdom
   96. Dominic P Kwiatkowski, Wellcome Trust Sanger Institute, Cambridge, United Kingdom

• 6,952

  Page views

• 1,430

  Downloads

• 185

  Citations

Article citation count generated by polling the highest count across the following sources: Scopus, Crossref, PubMed Central.