Decoding the evolution and transmissions of the novel pneumonia coronavirus (SARS-CoV-2) using the whole genomic data

Abstract

Background. The outbreak of COVID-19 started in mid-December 2019 in Wuhan, Central China. Up to February 18, 2020, SARS-CoV-2 has infected more than 70,000 people in China, and another 25 countries across five continents. In this study, we used 93 complete genomes of SARS-CoV-2 from the GISAID EpiFluTM database to decode the evolution and human-to-human transmissions of SARS-CoV-2 in the recent two months.

Methods. Alignment of coding-regions was conducted haplotype analyses using DnaSP. Substitution sites were analyzed in codon. Evolutionary analysis of haplotypes used NETWORK. Population size changes were estimated using both DnaSP and Arlequin. Expansion date of population size was calculated based on the expansion parameter tau (τ) using the formula t=τ/2u.

Findings. Eight coding-regions have 120 substitution sites, including 79 non-synonymous and 40 synonymous substitutions. Forty-two non-synonymous substitutions changed the biochemical property of amino acids. No evident combination was found. Fifty-eight haplotypes were classified as five groups, and 31 haplotypes were found in samples from both China and other countries, respectively. The rooted network suggested H13 and H35 to be ancestral haplotypes, and H1 (and its descendent haplotypes including all samples from the Hua Nan market) was derived H3 haplotype. Population size of SARS-CoV-2 were estimated to have a recent expansion on 6 January 2020, and an early expansion on 8 December 2019.

Interpretation. Genomic variations of SARS-CoV-2 are still low in comparisons with published genomes of SARS-CoV and MERS-CoV. Phyloepidemiologic analyses indicated the SARS-CoV-2 source at the Hua Nan market should be imported from other places. The crowded market boosted SARS-CoV-2 rapid circulations in the market and spread it to the whole city in early December 2019. Furthermore, phyloepidemiologic approaches have recovered specific direction of human-to-human transmissions, and the import sources of international infectious cases.

Full text


Decodingtheevolutionandtransmissionsofthenovel
pneumoniacoronavirus(SARS-CoV-2/HCoV-19)
usingwholegenomicdata
Wen-BinYu1,2,*,Guang-DaTang3,4,LiZhang5,RichardT.Corlett1,2
1Center for Integrative Conservation, Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences, Mengla, Yunnan
666303, China
2Center of Conservation Biology, Core Botanical Gardens, Chinese Academy of Sciences, Mengla, Yunnan 666303, China
3Henry Fok College of Biology and Agriculture, Shaoguan University, Shaoguan, Guangdong 512005, China
4College of Forestry and Landscape Architecture, South China Agricultural University, Guangzhou, Guangdong 510642, China
5Chinese Institute for Brain Research, Beijing 102206, China

ABSTRACT
Theoutbreak ofCOVID-19 startedin mid-December
2019 in Wuhan, China. Up to 29 February 2020,
SARS-CoV-2 (HCoV-19 / 2019-nCoV) had infected
morethan85 000people intheworld. Inthis study,
weused93completegenomesofSARS-CoV-2from
the GISAID EpiFluTM database to investigate the
evolution and human-to-human transmissions of
SARS-CoV-2inthefirsttwo monthsoftheoutbreak.
We constructed haplotypes of the SARS-CoV-2
genomes, performed phylogenomic analyses and
estimated the potential population size changes of
the virus. The date of population expansion was
calculatedbasedontheexpansionparametertau(τ)
using the formula t=τ/2u. A total of 120 substitution
siteswith119codons,including79non-synonymous
and 40 synonymous substitutions, were found in
eight coding-regions in the SARS-CoV-2 genomes.
Forty non-synonymous substitutions are potentially
associated with virus adaptation. No combinations
were detected. The 58 haplotypes (31 found in
samples from China and 31 from outside China)
wereidentifiedin 93viral genomesunderstudy and
couldbeclassifiedinto five groups. Byapplyingthe
reportedbatcoronavirusgenome(bat-RaTG13-CoV)
astheoutgroup, we foundthathaplotypesH13 and
H38 might be considered as ancestral haplotypes,
and later H1 was derived from the intermediate
haplotypeH3.ThepopulationsizeoftheSARS-CoV-
2 was estimated to have undergone a recent
expansion on 06 January 2020, and an early
expansion on 08 December 2019. Furthermore,
phyloepidemiologic approaches have recovered
specificdirections ofhuman-to-humantransmissions
and the potential sources for international infected
cases.
Keywords:COVID-19; HCoV-19; SARS-CoV-2;
Novel pneumonia outbreak; Human-to-human
transmission;Phyloepidemiology


Received:01 March 2020; Accepted: 27 April 2020; Online: 30 April
2020
Foundation items:This study was supported by grants from Ten
Thousand Talents Program of Yunnan for Top-notch Young Talents,
and the open research project of  “Cross-Cooperative Team” of the
Germplasm Bank of Wild Species, Kunming Institute of Botany,
ChineseAcademyofSciences
*Correspondingauthor,E-mail:yuwenbin@xtbg.ac.cn
DOI:10.24272/j.issn.2095-8137.2020.022
OpenAccess
This is an open-access article distributed under the terms of the
Creative Commons Attribution Non-Commercial License (http://
creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted
non-commercial use, distribution, and reproduction in any medium,
providedtheoriginalworkisproperlycited.
Copyright ©2020 Editorial Office of Zoological Research, Kunming
InstituteofZoology,ChineseAcademyofSciences


Received:01 March 2020; Accepted: 27 April 2020; Online: 30 April
2020
Foundation items:This study was supported by grants from Ten
Thousand Talents Program of Yunnan for Top-notch Young Talents,
and the open research project of  “Cross-Cooperative Team” of the
Germplasm Bank of Wild Species, Kunming Institute of Botany,
ChineseAcademyofSciences
*Correspondingauthor,E-mail:yuwenbin@xtbg.ac.cn
DOI:10.24272/j.issn.2095-8137.2020.022
OpenAccess
This is an open-access article distributed under the terms of the
Creative Commons Attribution Non-Commercial License (http://
creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted
non-commercial use, distribution, and reproduction in any medium,
providedtheoriginalworkisproperlycited.
Copyright ©2020 Editorial Office of Zoological Research, Kunming
InstituteofZoology,ChineseAcademyofSciences
ZOOLOGICALRESEARCH
SciencePress Zoological Research41(3):247−257,2020 247

INTRODUCTION
Betacoronaviruses are characterized by enveloped, positive-
sense, single-stranded RNA, and hosted in animals,
particularly mammals (Cui et al., 2019). Before December
2019,fourspecies/strains of Betacoronavirus,HKU1, MERS-
CoV, OC43, and SARS-CoV, had been reported to cause
severe human diseases (Cui et al., 2019). The fifth
species/strain,a novelbetacoronavirusSARS-CoV-2/HCoV-
19 / 2019-nCoV (Gorbalenya et al., 2020; Jiang et al., 2020)
causinghumanpneumonia(i.e.,COVID-19),wasfirstreported
inWuhan,Hubei,CentralChina(Wuetal.,2020a;Zhouetal.,
2020;Zhuetal.,2020).Upto29February2020,SARS-CoV-2
hadinfectedmorethan85000peopleinallprovinces/regions
of China, and another 59 countries/regions across Africa,
Asia, Europe, North America, Oceania, and South America
(Wikipedia, 2020). Because SARS-CoV-2 can transmit from
human to human (Li et al., 2020), the massive exodus of
people before the Chinese Spring Festival boosted the
infection frequencies, as predicted (Wu et al., 2020b). Daily
confirmedinfectioncases weremorethan 2000 between 30
January and 16 February, 2020, and the highest was more
than15100(Wikipedia,2020),almosttwice thetotal number
forSARS-CoV(Chan-Yeung&Xu,2003).
Asa memberof subgenusSarbecovirus, SARS-CoV-2has
beensuggestedtobeof batorigin (Luet al.,2020b;Zhou et
al.,2020),andmayhavebeentransmittedtohumansthrough
non-bat intermediate mammals (e.g., pangolins (Cyranoski,
2020;Lamet al.,2020;Wong etal.,2020; Xiaoet al., 2020;
Zhang et al., 2020b)). Medical information for the first 41
infected patients in Wuhan showed that 27 patients were
linkedtotheHuananSeafood WholesaleMarket(abbreviated
asHuananmarketinthetextbelow)(Huangetal.,2020;Liet
al., 2020), which sold living wild mammals. This suggests a
high possibility that SARS-CoV-2 originated in the market,
thenthe infectedpeopletransmittedittootherpeopleoutside
ofthemarket.However,this conclusionhas beenchallenged
becausethe firstidentifiedinfectedpersonand 12others had
no link to the Huanan Market. Some researchers have
thereforeargued thatthe HuananMarket wasnottheoriginal
and/or only source of SARS-CoV-2 transmission to humans
(Cohen, 2020). The market was closed on 01 January 2020,
making it very difficult to identify the intermediate animal
vectors of SARS-CoV-2. In the absence of information on
potential intermediary reservoirs, the origin and transmission
pattern of SARS-CoV-2 are still unresolved (Wong et al.,
2020).
Since the outbreak of COVID-19 was first identified in
Wuhan in mid-December 2019, the first infected individuals
identified in other provinces and regions of China, and other
countries,duringJanuary 2020,havebeen assumedto have
been infected in Wuhan or through contact with people from
Wuhan(Chanet al., 2020;Holshue et al.,2020; Phan etal.,
2020;Rothe etal.,2020).Inthis study,weused93genomes
of SARS-CoV-2 from the GISAID EpiFluTM database (Shu &
McCauley, 2017) (access date 12 February 2020) to decode
the evolution and transmissions of SARS-CoV-2 in the first
two months of its spread. Our aims were to: (1) characterize
genomicvariationsof SARS-CoV-2;(2)infer theevolutionary
relationships of the worldwide samples; and (3) deduce the
transmissionhistory ofSARS-CoV-2 withinWuhanandoutof
Wuhantotheworld.
MATERIALSANDMETHODS
To decode the evolutionary history of SARS-CoV-2, we
retrieved96 completegenomesfromGISAID(Supplementary
Table S1, access by 12 February 2020) (Shu & McCauley,
2017). The genome EPI_ISL_402131 (bat-RaTG13-CoV,
hereafter) from GISAID was also included as the outgroup,
because it is the closest sister betacoronavirus available to
SARS-CoV-2(Zhou etal.,2020).The97genomesequences
werealignedusingMAFFT(Katoh&Standley,2013),thenthe
alignmentwasmanuallycheckedusingGeneious(Biomatters,
New Zealand). In the alignment, we found that
EPI_ISL_404253 contains six ambiguous sites at variable
positions and EPI_ISL_407079 and EPI_ISL_408978 have
175  “N” and 1476  “N” bases, respectively, so these three
genomes were excluded in this study. In addition, four
genomes (EPI_ISL_407071, EPI_ISL_407894, EPI_ISL_
407896, and EPI_ISL_409067) have their own private
ambiguous sites, which were conservatively replaced by the
commonnucleotide atthatpositioninthealignment;Notably,
EPI_ISL_406592 (H15) and EPI_ISL_406595 (H17) had
excessive amounts of private variable sites, which were
possibly affected by sequencing errors. In the alignment, the
5'untranslated region (UTR)and 3' UTR regions contain
missingandambiguoussites,sotheseregionswereexcluded
inthefollowinganalyses.
ThealignmentwasthenimportedintoDnaSP(Rozasetal.,
2017) for haplotype analyses. Population size changes were
estimated based on a constant population size hypothesis
usingDnaSP,incombinationwith neutralitytests (Tajima’sD
and Fu’s Fs). We also used Arlequin (Excoffier & Lischer,
2010)totestthesuddenpopulationexpansionhypothesisand
tocalculatetheexpansionparametertau(τ),sincethesudden
populationexpansionwas not rejected.We used theformula
t=τ/2u(Rogers&Harpending,1992)toestimatethetimesince
expansion (in days). In the formula, u is the cumulative
substitution rate per year for the genome sequence, so we
used the formula u=μk to calculate it, where μ is the
substitution rate per site per year, and k is the genome
sequence length (29 358 bp for the coding sequence (CDS)
matrix). The substitution rate was set as 0.92×10−3 (95% CI,
0.33×10−3–1.46×10−3)substitution/site/yearbasedon themost
recentestimationforSARS-CoV-2(Rambaut,2020).Toadjust
the time, we used a mean value of the expansion time
calculated from the three substitution rates, i.e., 0.33×10−3,
0.92×10−3, and 1.46×10−3 substitution/site/year. In addition,
theexpansiondatewasestimatedbasedonthesamplingdate
fromhospitalizedpatients.Theestimateddateshouldbelater
thanthe“real”dateof massivehuman-to-humantransmission
events.
248www.zoores.ac.cn

Phylogeneticnetworksofthehaplotypecodingregionmatrix
and 120 substitution sites of SARS-CoV-2 (Supplementary
Datasets) were inferred using SplitsTree (Huson & Bryant,
2006).Amedian-joining networkofhaplotypeswas generated
by the NETWORK program (Bandelt et al., 1999, 2020) with
the reported bat coronavirus (bat-RaTG13-CoV, (Zhou et al.,
2020)) as the outgroup. Transversions were arbitrarily
weighted three times as high as transitions. Hypervariable
sites(ifnumberofmutations ≥5)wereweighted as1,and the
other sites were weighted as 10. Genetically, SARS-CoV-2
and bat-RaTG13-CoV, as well as SARS-CoV, have been
proposedas thesamespecies (Gorbalenyaetal.,2020),and
genome sequence identity between bat-RaTG13-CoV and
SARS-CoV-2 was 96.2%. We carefully used three datasets
(i.e.,fourcore substitutionsites, 120substitutionsites, and1
235 substitution sites, Supplementary Datasets) to evaluate
the relationship between bat-RaTG13-CoV and four
associated/central haplotypes of SARS-CoV-2 (H1, H3, H13,
and H38). Phylogenomic analyses of haplotypes were
performedusingIQ-TREE (Minhetal., 2020).We conducted
likelihood mapping and SH-like approximate likelihood ratio
tests to assess the phylogenetic information and branch
supports,respectively.
RESULTSANDDISCUSSION
GenomicvariationsofSARS-CoV-2
GenomesizeofSARS-CoV-2variedfrom29782bpto29903
bp.Thealignedmatrixwas29910bpinlength,including140
variable sites. The CDS regions contained 120 substitution
sites(Supplementary FigureS1),whichwereclassifiedas58
haplotypes (Supplementary Table S2). Nucleotide diversity
(Pi) was 0.15×10−3±0.02×10−3 (standard deviation, SD,
hereafter).Haplotypediversity(Hd)was0.953±0.016(SD)and
varianceofHdwas0.26×10−3.
There were 120 substitution sites found in eight coding
sequence (CDS) regions of SARS-CoV-2 (Figure1,
Supplementary Table S2), including 79 transitions (65.83%)
and 41 transversions (34.17%). A chi-squared test showed
thatthedistributionofsubstitutionsitesacrossCDSregionsin
the genome was even (χ2=1.958, df=9, P=0.99). Substitution
sites occurring at the 1st to 3rd frame positions were 27
(25.55%),44(40.0%),and49(44.55%),respectively.The120
substitutionsiteswere associated with119 codons, including
79 non-synonymous (65.83%) and 40 synonymous (33.61%)
substitutions. Forty non-synonymous substitutions (50.63%)
changed the biochemical properties of the amino acid (AA),
andare thereforepotentially associatedwithvirusadaptation.
The current samplings showed that the H1 haplotype has
been found in 19 patients, but most haplotypes were just
sequenced once, suggesting that the haplotype H1 was
rapidly circulated at an early stage of human-to-human
transmissions(Figure2,SupplementaryTableS1).
Incomparisonswith publishedgenomesofSARS-CoV (Luk
et al., 2019) and MERS-CoV (Cotten et al., 2013), genomic
variations of SARS-CoV-2 are still low, without evident

Figure1Summaryinformationfor 120substitution sites crossingeight codingsequence regions inthe alignedSARS-CoV-2genomic
sequences

Zoological Research41(3):247−257,2020 249

recombination sites/regions (Rm=2, P=1.0) at this time.
According to the collection dates of the sequenced samples,
haplotypesH1andH3were foundintwo samplesatintervals
of more than 30 days, and multiple samples over 20 days
(Figure2,Supplementary Table S1).Although the incubation
periodcanbe over24days, there wasonly one caseof this
out of 1 099 observations (Guan et al., 2020). Estimation of
the substitution rates using 90 genomes of SARS-CoV-2
(Rambaut, 2020) showed that the rate for SARS-CoV-2 was
closeto orlower thanthe ratesfor MERS-CoV(Cotten etal.,
2014;Dudasetal.,2018)andSARS-CoV(Zhaoet al.,2004).
Due to the mild symptoms and low mortality (Yang et al.,
2020; Zhang et al., 2020a), the immune systems of the
infected humans may provide a suitable environment for
propagation of SARS-CoV-2 (Andersen et al., 2020). SARS-
CoV-2 is highly infectious (Yang et al., 2020) and is able to
infecthumansnotonlythroughthemucousmembranesofthe
noseand mouth,butalsousethemucousmembranesinthe
eyes (Lu et al., 2020a), which may boost regional circulation
and large-scale spread. Some large mutations may have
occurredinWuhan or otherregions, but thestrict quarantine
policy over China since 23 January 2020 may have reduced
thecirculationandspreadingofsomemutants.
Ofthe93genomesofSARS-CoV-2,39(41.93%)werefrom
infected patients in 11 countries outside China and encoded
31 haplotypes (Hd=0.987±0.009 (SD), Pi=0.16×10–3±
0.01×10–3), with 27 nationally/regionally private haplotypes.
The 54 genomes (58.07%) from China also encoded 31
haplotypes (Hd=0.906±0.001 (SD), Pi=0.14×10–3±0.03×10–3).
AproportionZ-testshowedsignificantdifferencesinhaplotype
diversity of samples between China and other countries
(χ2=4.024,df=1,P<0.05).Thehighhaplotypediversityfoundin
samples from other countries may be because the sampling
datesweremostlyafter22January2020,whilethoseinChina
were before this date (Supplementary Table S1 and Figure
S2). In addition, the low level of radiation exposure on long-
distance international flights (Bottollier-Depois et al., 2000)
mayhaveacceleratedmutationrates ofSARS-CoV-2 (Shibai
etal.,2017).
PopulationsizeexpansionofSARS-CoV-2
We used a variety of parameters to estimate the population
dynamicsofSARS-CoV-2.ConstantpopulationsizeofSARS-
CoV-2 was rejected (Ramos-Onsins and Rozas’s R2=0.025,
P<0.001;Raggedness r=0.011,P<0.05) usingDnaSP(Rozas
et al., 2017) (also see Supplementary Figure S3), while both
Fu's test (Fs=–67.681.964, P<0.001) and Tajima's D test
(D=–2.701, P<0.001) indicated that the population size of
SARS-CoV-2 was rapidly increasing. Mismatch distribution
analysis using Arlequin (Excoffier & Lischer, 2010) strongly
supported that the population of SARS-CoV-2 underwent
sudden expansion (τ=2.887, Sum of Squared deviation,

Figure2GenomichaplotypesofSARS-CoV-2changesbetweenthecollectiondatesofsamples
Theconfirmedsamples fromthe HuananSeafood WholesaleMarket areindicated usingredcircles, anda confirmedsample withno linktothe
marketisindicatedusingabluecircle.

250www.zoores.ac.cn

SSD=0.541×10–3, P=0.88, Harpending's Raggedness index,
R=0.010,P=0.88).The calculatedexpansion was28.72 days
(95% Confident Interval: 12.29–54.36 days) ago. Of the 93
genomes, the latest one was sampled on 03 February 2020,
so the estimated expansion date was on 06 January 2020
(95%CI:11December2019–22January2020),whichmaybe
relatedtotheNewYearholiday.Before06January2020,129
patientswereidentifiedasSARS-CoV-2 infectedthroughfield
investigations(Lietal.,2020).Of22genomes(17.05%of129
patients) sequenced before 06 January 2020 in Wuhan,
China, 13 haplotypes (22.41% of 58 haplotypes) were
recovered, which were H1 and its derived descendant
haplotypes, and H3 (Figures 2 and 3A). Coincidentally, the
China CDC (Chinese Center for Disease Control and
Prevention)startedtoactivatea Level-2emergencyresponse
on 06 January 2020 (Li et al., 2020). The China CDC’s
emergency response greatly reduced public activities and
travel,andmighthavereducedthelocalcirculationandlarge-
scalespreadinthefollowingweeksofJanuary.
Furthermore, mismatch distribution analysis of the 22
genomes before 06 January 2020 also showed a sudden
population expansion of SARS-CoV-2 at an earlier stage of
transmission(τ=2.818,SSD=0.010,P=0.41,R=0.046,P=0.57,
Tajima’sD=–2.241,P<0.001; Fu'sFs=–7.834,P<0.001). This
earlierpopulationexpansiontimewasestimatedat28.38days
(95% CI: 12.00–54.36 days) before 05 January 2020, which
wasthelatest samplingdate ofthe22 genomes.This earlier
expansion date was thus estimated to have occurred on 08
December 2019 (95% CI: 13 November 2019–26 December
2019), when there was only one infected patient officially
reported (Huang et al., 2020; Li et al., 2020). This suggests
thatSARS-CoV-2mighthavealreadycirculatedwidelyamong
humansinWuhanbeforeDecember2019,probablybeginning
inmidtolateNovember(Rambaut,2020).
EvolutionaryrelationshipsofSARS-CoV-2haplotypes
Phylogenetic networks showed that the 58 haplotypes were
clusteredintotwomainclades (Figure4).Clade Iincluded19
haplotypesandCladeIIincluded39haplotypes.Theoutgroup
bat-RaTG13-CoV was connected to Clade I, supposed to be
anancestralcladeforCladeII.ThelongbranchesofH15and
H17 correspond to an excessive amount of mutations, which
are possibly affected by sequencing errors, but this is still to
be determined. Three different datasets were used to infer
evolutionarynetworks,which consistentlysupported H13 and
H38as thepotentially ancestralhaplotypes,i.e.,theoutgroup
bat-RaTG13-CoVcouldconnecttobothH13andH38,orH38
alone,or throughamediumvectormv1 (anintermediate host
orthefirstinfectedhumans) connectedto bothH13 andH38
by single mutations at positions 18067 (S, synonymous
substitution) and/or 29102 (S), referring to the numbering of
thealignmentlength 29 910bp (Figure 5).Five main groups
can be recognized in the network using the dataset of 120
substitutionsites(Figure3A).TheH1,H3,andH13werethree
core haplotypes, so that Groups A–C were recognized using
them as the central (i.e., ancestral super-spreader)
haplotypes. Groups D and E were recognized based on two
new super-spreader haplotypes, H56 and a medium vector
mv2, which was a hypothesized (often ancestral) haplotype
notsampledinthecurrentsamples.Thesetwogroupscanbe
alsotreatedassubgroupsofGroupC. Moreover,the SH-like
approximate likelihood ratio test further enhanced the

Figure3Evolutionaryrelationshipandgeographicaldistributionof58haplotypesofSARS-CoV-2(A,B)
Proposed evolutionary paths (C) of haplotypes and possible transmission and spreading routes (D) are also inferred based on evolutionary
analysesandepidemiologicresearch.Samplesizesofhaplotypesandregionsareannotatedinthecircles.

Zoological Research41(3):247−257,2020 251

phylogenetic relationship retrieved from 58 haplotypes, that
eitherH13or H38(withH45) appearedin the basallineages
(Supplementary Figure S4), although it was difficult to
distinguish which one, H13 or H38, originated earlier

Figure4Phylogenetic networks of 58 haplotypes of SARS-CoV-2 with the outgroup bat-RaTG13-CoV using the whole coding region
matrix(A)and120substitutionsites(B),referringtoallvariablesitesofcodingregionsinSARS-CoV-2haplotypes
Thebottomscalebarsrepresentthenumberofsubstitutions persite.Theopenbox onbranchesofbat-RaTG13-CoV(A),H15(A,B),andH17(A,
B)indicatingalongbranchwasclipped.

252www.zoores.ac.cn

(SupplementaryFigureS5).
Inthenetwork,foursatellitehaplotypes andH35connected
toH13(GroupA),andnine satellitehaplotypesandH38+H45
and H50 connected to H3 (Group B). The connections
between the H3 and H1 are two mutations at positions 8789
(S)and28151(NS,non-synonymous substitution)(Figure 5),
thelattermutationchangedbothresiduesandthebiochemical
properties of the AA. This biochemical change may be
associated with the infectivity of SARS-CoV-2. The H1
haplotype,the mostabundant, included19 samples,while 26
satellite haplotypes and H40+(H43 and H47) haplotypes are
directlyderivedfromH1(GroupC).Moreover,fivehaplotypes
ofGroupDandfourhaplotypesofGroupEwerealsoderived
fromH1.
The Huanan Seafood Wholesale Market boosted human-
to-humantransmissionatanearlystage
Phylogenetic networks showed that bat-RaTG13-CoV was
nested with Group B in Clade I, and Clade II tends to be
derived from Clade I, i.e., H1 and its descendant haplotypes
werenewmutantsfromanancestralhaplotypeinCladeI.The
rootednetworksuggested two potentialevolutionary paths of
availablehaplotypesthat can befrom H13 throughH3 to H1
andH38,or fromH38throughH3toH1andH13(Figure3C).
Both scenarios suggested that H3 might be the ancestral
haplotypeofH1.H13was onlyrecoveredfrom fiveShenzhen
(Guangdong Province) samples, including patient 2 of the
familial cluster (Chan et al., 2020). Two derived haplotypes
were also only found in Shenzhen of Guangdong Province
(H14 from the grandson of patient 2), and the other three
haplotypeswerefoundinthree samplesfrom Japanand one
sample from Arizona in the United States (Figure 3).
According to an epidemiological study, the Shenzhen family
couldhave beeninfectedduring theirvisittoWuhan(Chan et
al.,2020). ThissuggeststhatH13mighthaveoriginated from
Wuhan.Genetically, haplotypesofGroupAhavelinkstoonly
Wuhan haplotype H3 (only EPI_ISL_406801). It is possible
thatH13was newlyderived fromH3 (Figure3C)and didnot
spreadinWuhan,orthatthreerepatriatedJapanesemightbe
infected by an unknown source of H13 in Wuhan, China or
somewhere else (The Asahi Shimbun, 2020), or that no
samples have been sequenced yet. H38 has three genomes
from the same patient (Supplementary Table S1), who was
the first identified infected patient in the United States
(Holshueetal., 2020).Thispatient mighthave been infected
while visiting his family in Wuhan, China, or was infected in
some other place. The original source of H38 can be
explainedasthatofH13, whichcan bealso derivedfromH3
(Figure 3C), and the derived H45 was from a Chongqing
patient who was reported as working in Wuhan and had no
linktotheHuananMarket.
TheH3haplotypehasonly onesamplefrom Wuhan,which
was not linked to the Huanan Market (Lu et al., 2020b), and
the other samples in this group were from outside of Wuhan
(Figure 3A). Noteworthily, all the samples from the Market
belonged to H1 or its derived haplotypes (H2, H8-H12, see
Figure 2 and Supplementary Table S1), indicating that there
werecirculatedinfections withinthe marketin theshortterm.
Other researchers have argued that the source of the
coronavirusintheMarketshouldbeimportedfromelsewhere,
oratleast itshould be notthe singlesourceof SARS-CoV-2

Figure5The inferred relationships between the outgroup bat-RaTG13-CoV and four associated/central node haplotypes (H1, H3, H13,
andH38)ofSARS-CoV-2usingthreedatasets
The dataset of four core substitution sites is the four variable sites shared among bat-RaTG13-CoV and the four central node haplotypes. The
dataset of 120 substitution sites refers to all variable sites of coding regions in SARS-CoV-2 haplotypes. The dataset of 1235 substitution sites
referstoallvariablesitesofcodingregionsamongbat-RaTG13-CoVandSARS-CoV-2haplotypes.

Zoological Research41(3):247−257,2020 253

(Cohen, 2020). In this study, evolutionary relationships
indicated that H1 and its descendant haplotypes from the
Marketshould bederivedfromH3(Figures 3,4). H3mutated
to the H1 by two substitutions, and none of the currently
available Market samples encoded H3, suggesting that H3
mighthaveoriginatedandspreadoutsideoftheMarketbefore
anearlystageofpopulation expansion.Thenon-synonymous
mutation from H3 to H1 might have enhanced the
infectiousness of SARS-CoV-2, and a functional
characterization should be performed to confirm this
speculation.ItispossiblethatSARS-CoV-2intheMarket had
been transmitted from other places (Figure 3D), or at least,
thattheMarket didnothostthe originalsourceofSARS-CoV-
2(Cohen,2020).Asthefirstidentifiedinfectedpatientshadno
link to the Market (Huang et al., 2020), it is possible that
infectedhumanstransmittedtheH1haplotypeofSARS-CoV-2
to workers or sellers in the market, after which it rapidly
circulatedthereduetoits specialsurroundings. Thecrowded
market boosted SARS-CoV-2 transmissions to buyers and
spread it to the whole city in early December 2019,
corresponding to the estimated population expansion time.
Due to insufficient sampling from Wuhan in the currently
availablesamples,it isnot clearwhetherH3 neverappeared
intheMarket, orH1 wasquicklyderived fromH3 toadaptin
theMarket.
Regionalandworldwidecirculationandspread
Of the 54 genomes from patients in China, Chongqing (3
samples), Guangdong (18), Hubei (22), Taiwan (2), and
Zhejiang(4)have morethan twosamples,and theother five
provinces have one sample. Hubei (Wuhan) samples dated
from 24 December 2019 to 05 January 2020 encoded 13
haplotypes, belonging to Groups C (H1 and 11 satellite
haplotypes) and B (only H3). These relationships indicated a
rapid transmission and circulation of SARS-CoV-2 in Wuhan
at an early stage of human-to-human transmissions. H1 (no
satellite haplotypes) and H3 are the ancestors of haplotypes
outside of Wuhan/Hubei because most of early confirmed
patients might have history in Wuhan or Hubei. Eighteen
Guangdong samples, collected from 10–23 January 2020,
encoded 15 haplotypes, belonging to Groups A, C, and E,
showing that there were multiple sources imported into
Guangdong.Threehaplotypes(H14,H15,andH17)mayhave
evolvedlocally,indicatingthathuman-to-humantransmissions
happened when SARS-CoV-2 initially spread to Shenzhen in
Guangdong Province (Chan et al., 2020). Two samples from
Taiwan Province, China, encoded H3 and H24 in Groups B
and D, respectively, and three samples from Chongqing
encodedH1,H40, andH45 inGroups BandC, respectively.
There were two sources imported into these two provinces.
Four Zhejiang samples encoded H1 and H24 in Group C,
which might be only imported from the source of the H1
haplotype.
The samples outside China encoded 31 haplotypes
belongingtoGroupsA–E.Ofthese,27haplotypesareprivate
by regional samplings, only two samples from Thailand were
theH1 haplotype,one eachfrom Australiaand Belgiumwere
the H3 haplotype, one from the United States was the H19
haplotype, and one from Singapore was the H40 haplotype.
Twelvesamples,encoding10haplotypes,werefrom patients
infive countriesin Asia.Sixhaplotypeslinkedto H1,and two
each linked to H3 and H1, respectively, indicating the 12
patientswereinfected bydifferent sources. Human-to-human
transmissionsmayhave happened frompatients with H53to
H52 haplotypes in Tokyo, Japan. Five Australian samples,
encoding six haplotypes in Groups B, C, and D, were from
patientsofthree states. Patientswith H3, H25and H26, and
withH55were in GroupsB and C,respectively, and human-
to-human transmission might have happened from the
patients with H25 to H26, who were in a same tour group in
Queensland (AAP reporters, 2020). The connection between
thepatientswith H56and H27 isnot clear.Onepossibility is
that there was an intermediary spreader with H56, who also
transmitted SARS-CoV-2 to other patients in France, the
UnitedStates,andTaiwanProvinceofChina.EightEuropean
samples, encoding seven haplotypes, were from patients in
four countries. The patients in England were reported as a
household transmission from H28 to H29 (Lillie et al., 2020).
The patients in France may have been infected by three
differentsources,i.e.,H44waslinkedtoH1,H43mightlinkto
H40(inChongqing, Singaporeorsomewhere else),and H30
might link to an intermediary spreader with H56. Of the 13
genomes from the United States, three were from the same
patient in Washington encoding the same haplotype H38,
while the other ten samples encoded eight haplotypes,
covering all five groups (Figure 3A, B), so the sources of
infectionsarecomplicated. Thereisnoevidenceofhuman-to-
humantransmissionintheUnitedStatesfromthese11cases.
Toclarifytheexactoriginsofthesehaplotypesoutside China,
we need more epidemiological investigative efforts and more
SARS-CoV-2genomicdatafrompatientsattheearlystageof
transmissions.
Phylogenetic approaches provide insights into the
epidemiologyofSARS-CoV-2
Epidemiological study of SARS-CoV-2 using traditional
approachesis verydifficult,because itwasnotidentifiedas a
newcoronavirusuntil29December,andsomeinfectedpeople
withmildsymptomsorwithoutsymptoms(Heymann&Shindo,
2020;Rotheet al., 2020;Wu & McGoogan,2020) may have
been overlooked in late November and early December.
Evolutionary analyses suggested that the source of the H1
haplotype in the Huanan Market was imported from
elsewhere, as has been suggested by other researchers
(Cohen, 2020). The rooted network suggested that H13 and
H38 should be ancestral haplotypes that connected to the
outgroup bat-RaTG13-CoV through a hypothesized
intermediate haplotype (Figure 5). The most common
ancestral haplotype was missed because the currently
available samples do not include the first identified infected
patientand otherpatientsfromearlyDecember, andbecause
oftherelativelyhighmutationrateoftheviralgenome.Ifthere
254www.zoores.ac.cn

areanyfrozensamplesfromthosepatients,itwouldbe worth
doing genomic sequencing for phyloepidemiologic study to
help to locate the birthplace of SARS-CoV-2. Meanwhile, we
expect that the H13 and H38 haplotypes might be found in
some samples from infected patients in Wuhan or in other
places across the world if more samples are sequenced in
future. This will be very helpful in the search for the original
sourcesofSARS-CoV-2, becauseboth H13and H38tendto
beancestralhaplotypes.
The evolutionary network of haplotypes can be used to
recover the directions of human-to-human transmissions at
the local scale and spread at the larger scale. The central
haplotype can be considered as the super-spreader
haplotype, and the tip haplotype is the most recent
descendant, similar to the definition and use of mtDNA
haplogroupsintracinghumandemographichistory(Yaoetal.,
2002). The transmission direction can be identified using the
connectioninformationoftipsandbranches.Forexample,the
confirmed patients from the Huanan Market shared the
common ancestral haplotype H1, indicating they might be
infectedfromacommonsource,whomayhavebeenasuper-
spreader in the market. This approach has recovered
potentiallyspecificdirectionsofhuman-to-humantransmission
in the Shenzhen family (H13 → H14), the Queensland tour
group(H25→H26),theEnglandfamily(H28→H29),andthe
Japanese (H53 → H52). It is possible that some infections
couldlinktoWuhanorHubeidirectlyorindirectly,becausethe
patientsclaimedconnectionstoWuhanorHubei,butforsome
of them it is not clear exactly where they were infected. We
suspectthatthereweresuper-spreadersmediatingthespread
ofSARS-CoV-2attheearlystageoftransmissions.
Our findings showed that SARS-CoV-2 has not had
legitimate recombination. Thus, the haplotype-based
phyloepidemiologic analyses provide a powerful way to
understand the evolution of SARS-CoV-2 at the very early
stageoftransmissionwhenreverse mutationsandillegitimate
recombination are rare. In our analysis, recombination is
rejectedbuttheoutgroupbat-RaTG13-CoV isrelativelyhighly
divergedfromSARS-CoV-2haplotypes,which mayaffect the
phyloepidemiologic analyses. Based on the estimated
mutation rate of current SARS-CoV-2 viruses, the reverse
mutationsshouldbe6×10−3(0.92×10−3×0.92×10−3persiteper
year×29358sites×2/12year),with aneglectableinfluence on
ourresult.Butourobservationsleaveoneimportantquestion:
whyareancestralhaplotypes,likeH13andH38,lessfrequent
than H1? It is highly possible that H1 acquired adaptive
mutations,such asNS ofsite 28151,from H3orH13(and/or
H38),evolved inanindependentcirculationafter theyjumped
into intermediate hosts or directly transmitted to humans,
which should be investigated in future studies if more early
genomedatasetsareavailable.The exactoriginal sourcesof
H13andH38willstayasunsolvedmysteriesiftheearlystage
sampleswerenotpreserved.
Anearlyversionofourmanuscriptwas postedat ChinaXiv
(DOI: 10.12074/202002.00033) on 19 February 2020. Since
then,therehavebeenmanynewsstories stemmingfrom our
manuscript with a biased interpretation of the results. This is
beyondour expectation.During thereviewofthismanuscript,
thereweresomereportsofanalysesofSARS-CoV-2genomic
variationsbasedon alargersample size(e.g.,Forster etal.,
2020;Tangetal.,2020),whichshowedasimilarphylogenetic
pattern as we present here. We expect more data-mining of
the increasing number of SARS-CoV-2 genomes will provide
updatedinsightsintotheoriginandtransmissionofthisvirus.
SUPPLEMENTARYDATA
Supplementarydatatothisarticlecanbefoundonline.
COMPETINGINTERESTS
Theauthorsdeclarethattheyhavenocompetinginterests.
AUTHORS’CONTRIBUTIONS
W.B.Y.conceivedtheresearch,analyzedthedata,interpreted
theresults,andwrotethedraftmanuscript;W.B.Y.andG.D.T.
collecteddata.Allauthorsreadandapprovedthefinalversion
ofthemanuscript.
ACKNOWLEDGEMENTS
We are grateful to scientists and researchers for depositing
whole genomic sequences of Novel Pneumonia Coronavirus
(SARS-CoV-2/HCoV-19/2019-nCoV)attheGlobalInitiative
on Sharing All Influenza Data (GISAID) EpiFlu™; to GISAID
database for allowing us to access the sequences for non-
commercialscientificresearch; and totwo reviewers fortheir
valuable comments and suggestions. This study was
supported by grants from Ten Thousand Talents Program of
Yunnanfor Top‐notchYoung Talents,andtheopenresearch
projectof “Cross-CooperativeTeam”oftheGermplasmBank
of Wild Species, Kunming Institute of Botany, Chinese
AcademyofSciences.
REFERENCES

AAP Reporters. 2020 (2020-01-30). Coronavirus outbreak: second case
confirmed in Queensland. https://7news.com.au/lifestyle/health-wellbeing/
qld-coronavirus-case-remains-in-isolation-c-671500.

Andersen KG, Rambaut A, Lipkin WI, Holmes EC, Garry RF. 2020. The
proximaloriginofSARS-CoV-2.Nature Medicine,26(4):450−452.

BandeltHJ, Forster P,Röhl A.1999. Median-joiningnetworks forinferring
intraspecificphylogenies.Molecular Biology and Evolution,16(1):37−48.

Bandelt HJ, Forster P, Röhl A. 2020 (2020-02-01). Free phylogenetic
networksoftware.https://www.fluxus-engineering.com/sharenet.htm.

Bottollier-DepoisJF, ChauQ, BouissetP, KerlauG, PlawinskiL, Lebaron-
Jacobs L. 2000. Assessing exposure to cosmic radiation during long-haul
flights.Radiation Research,153(5):526−532.

ChanJFW,Yuan SF,Kok KH,To KKW,Chu H,Yang J,Xing FF,LiuJL,
YipCCY, PoonRWS,TsoiHW,LoSKF,Chan KH,PoonVKM, ChanWM,
IpJD,Cai JP,Cheng VCC,Chen H,Hui CKM,Yuen KY.2020. Afamilial
clusterofpneumonia associatedwiththe2019 novelcoronavirusindicating
person-to-person transmission: a study of a family cluster. The Lancet,
Zoological Research41(3):247−257,2020 255

395(10223):514−523.
Chan-Yeung M, Xu RH. 2003. SARS: epidemiology. Respirology, 8(S1):
S9−S14.

Cohen J. 2020. Wuhan seafood market may not be source of novel virus
spreadingglobally.Science,doi:10.1126/science.abb0611.

CottenM, WatsonSJ,KellamP,Al-RabeeahAA,Makhdoom HQ,AssiriA,
Al-TawfiqJA, AlhakeemRF, MadaniH, AlRabiahFA,HajjarSA,Al-Nassir
WN, Albarrak A, Flemban H, Balkhy HH, Alsubaie S, Palser AL, Gall A,
Bashford-RogersR,RambautA,ZumlaAI,MemishZA.2013.Transmission
andevolutionof theMiddleEastrespiratorysyndromecoronavirusinSaudi
Arabia:adescriptivegenomicstudy.The Lancet,382(9909):1993−2002.
Cotten M, Watson SJ, Zumla AI, Makhdoom HQ, Palser AL, Ong SH, Al
Rabeeah AA, Alhakeem RF, Assiri A, Al-Tawfiq JA, Albarrak A, Barry M,
ShiblA,AlrabiahFA,HajjarS, BalkhyHH,FlembanH,RambautA,Kellam
P,Memish ZA.2014. Spread,circulation, andevolution ofthe middleeast
respiratorysyndromeCoronavirus.mBio,5(1):e01062−13.

CuiJ,LiF, ShiZL.2019.Originandevolution ofpathogeniccoronaviruses.
Nature Reviews Microbiology,17(3):181−192.

CyranoskiD.2020.DidpangolinsspreadtheChinacoronavirustopeople?.
Nature,doi:10.1038/d41586-020-00364-2.
DudasG,Carvalho LM,RambautA,Bedford T.2018.MERS-CoVspillover
atthecamel-humaninterface.eLife,7:e31257.

Excoffier L, Lischer HEL. 2010. Arlequin suite ver 3.5: a new series of
programs to perform population genetics analyses under Linux and
Windows.Molecular Ecology Resources,10(3):564−567.
Forster P, Forster L, Renfrew C, Forster M. 2020. Phylogenetic network
analysis of SARS-CoV-2 genomes. Proceedings of the National Academy
of Sciences of the United States of America, doi: 10.1073/pnas.
2004999117.

Gorbalenya AE, Baker SC, Baric RS, de Groot RJ, Drosten C, Gulyaeva
AA, Haagmans BL, Lauber C, Leontovich AM, Neuman BW, Penzar D,
Perlman S, Poon LLM, Samborskiy DV, Sidorov IA, Sola I, Ziebuhr J,
CoronaviridaeStudyGroup oftheInternationalCommitteeon Taxonomyof
V. 2020. The species severe acute respiratory syndrome-related
coronavirus: classifying 2019-nCoV and naming it SARS-CoV-2. Nature
Microbiology,5(4):536−544.

GuanWJ, NiZY, HuY,Liang WH,Ou CQ,HeJX, LiuL, ShanH, LeiCL,
Hui DSC, Du B, Li LJ, Zeng G, Yuen KY, Chen RC, Tang CL, Wang T,
ChenPY, XiangJ, LiSY, WangJL,LiangZJ,PengYX,WeiL,LiuY,Hu
YH,PengP, WangJM,LiuJY,Chen Z,LiG,ZhengZJ, QiuSQ,LuoJ,Ye
CJ, Zhu SY, Zhong NS. 2020. Clinical characteristics of coronavirus
disease 2019 in China. The New England Journal of Medicine, doi:
10.1056/NEJMoa2002032.

HeymannDL, Shindo N.2020. COVID-19:what isnext forpublic health?.
The Lancet,395(10224):542−545.

HolshueML,DeBolt C,LindquistS,LofyKH, WiesmanJ,BruceH,Spitters
C,EricsonK,WilkersonS,TuralA,DiazG,CohnA,Fox L,PatelA,Gerber
SI,KimL,TongSX,LuXY,LindstromS,PallanschMA,Weldon WC,Biggs
HM,Uyeki TM,PillaiSK. 2020.Firstcase of2019novel coronavirusinthe
UnitedStates.The New England Journal of Medicine,382(10):929−936.

HuangCL,WangYM,LiXW,RenLL,ZhaoJP,HuY,ZhangL,FanGH,Xu
JY,Gu XY,Cheng ZS,YuT, XiaJA, WeiY,Wu WJ,Xie XL,YinW, LiH,
LiuM,XiaoY, GaoH,GuoL,XieJG, WangGF,JiangRM,GaoZC,JinQ,
Wang JW, Cao B. 2020. Clinical features of patients infected with 2019
novelcoronavirusinWuhan,China.The Lancet,395(10223):497−506.
Huson DH, Bryant D. 2006. Application of phylogenetic networks in
evolutionarystudies.Molecular Biology and Evolution,23(2):254−267.

Jiang S, Shi Z, Shu Y, Song J, Gao GF, Tan W, Guo D. 2020. A distinct
nameisneededforthenewcoronavirus.The Lancet,395(10228):949.

KatohK,StandleyDM.2013.MAFFTmultiplesequencealignmentsoftware
version 7: improvements in performance and usability. Molecular Biology
and Evolution,30(4):772−780.

LamTTY,ShumMHH,ZhuHC,TongYG,Ni XB,LiaoYS,WeiW, Cheung
WYM, Li WJ, Li LF, Leung GM, Holmes EC, Hu YL, Guan Y. 2020.
Identification of 2019-nCoV related coronaviruses in Malayan pangolins in
southernChina.bioRxiv,doi:10.1101/2020.02.13.945485.

Li Q, Guan X, Wu P, Wang X, Zhou L, Tong Y, Ren R, Leung KSM, Lau
EHY,WongJY,XingX,Xiang N,WuY,LiC,ChenQ, LiD,LiuT, ZhaoJ,
LiuM,TuW,Chen C,JinL,YangR,WangQ,ZhouS,WangR,LiuH,Luo
Y,LiuY,ShaoG,Li H,TaoZ,Yang Y,DengZ,Liu B,MaZ,ZhangY,Shi
G, Lam TTY, Wu JT, Gao GF, Cowling BJ, Yang B, Leung GM, Feng Z.
2020.Early transmissiondynamicsin Wuhan,China, ofnovelcoronavirus-
infected pneumonia. The New England Journal of Medicine, 382(13):
1199−1207.

Lillie PJ, Samson A, Li A, Adams K, Capstick R, Barlow GD, Easom N,
Hamilton E, Moss PJ, Evans A, Ivan M, Phe Incident Team, Taha Y,
Duncan CJA, Schmid ML, the Airborne Hcid Network. 2020. Novel
coronavirusdisease(Covid-19):Thefirsttwo patientsintheUKwithperson
topersontransmission.Journal of Infection,80(5):578−606.

LuCW,LiuXF,JiaZF. 2020a.2019-nCoVtransmissionthrough theocular
surfacemustnotbeignored.The Lancet,395(10224):e39.

LuRJ, ZhaoX, LiJ, NiuPH,YangB,WuHL,WangWL,SongH,Huang
BY,Zhu N,BiYH, MaXJ, ZhanFX,Wang L,HuT,ZhouH, HuZH,Zhou
WM,ZhaoL,ChenJ,MengY, WangJ,LinY,YuanJY,XieZH,MaJM,Liu
WJ,Wang DY,XuWB, HolmesEC, GaoGF, WuGZ, ChenWJ, ShiWF,
TanWJ. 2020b.Genomiccharacterisationandepidemiologyof2019 novel
coronavirus:implicationsfor virusoriginsandreceptorbinding. The Lancet,
395(10224):565−574.
LukHKH,LiX,FungJ,LauSKP, WooPCY.2019.Molecularepidemiology,
evolution and phylogeny of SARS coronavirus. Infection, Genetics and
Evolution,71:21−30.
Minh BQ, Schmidt H, Chernomor O, Schrempf D, Woodhams M, von
Haeseler A, Lanfear R. 2020. IQ-TREE 2: new models and efficient
methods for phylogenetic inference in the genomic era. Molecular Biology
and Evolution,doi:10.1093/molbev/msaa015.
PhanLT,Nguyen TV,LuongQC,Nguyen TV,NguyenHT,Le HQ,Nguyen
TT, Cao TM, Pham QD. 2020. Importation and human-to-human
transmissionof anovel coronavirusin Vietnam.The New England Journal
of Medicine,382(9):872−874.

Rambaut A. 2020 (2020-02-12). Phylodynamic analysis|129 genomes |24
Feb 2020. http://virological.org/t/phylodynamic-analysis-90-genomes-12-
feb-2020/356.
Rogers AR, Harpending H. 1992. Population growth makes waves in the
distributionofpairwisegeneticdifferences.Molecular Biology and Evolution,
9(3):552−569.

Rothe C, Schunk M, Sothmann P, Bretzel G, Froeschl G, Wallrauch C,
Zimmer T, Thiel V, Janke C, Guggemos W, Seilmaier M, Drosten C,
Vollmar P, Zwirglmaier K, Zange S, Wölfel R, Hoelscher M. 2020.
256www.zoores.ac.cn

Transmission of 2019-nCoV infection from an asymptomatic contact in
Germany.The New England Journal of Medicine,382(10):970−971.
RozasJ, Ferrer-MataA, Sánchez-DelBarrioJC, Guirao-RicoS, LibradoP,
Ramos-Onsins SE, Sánchez-Gracia A. 2017. DnaSP 6: DNA sequence
polymorphismanalysisoflargedatasets. Molecular Biology and Evolution,
34(12):3299−3302.
Shibai A, Takahashi Y, Ishizawa Y, Motooka D, Nakamura S, Ying BW,
Tsuru S. 2017. Mutation accumulation under UV radiation in Escherichia
coli.Scientific Reports,7:14531.
ShuYL,McCauleyJ.2017.GISAID:Globalinitiativeonsharingallinfluenza
data-fromvisiontoreality.Eurosurveillance,22(13):30494.
Tang X, Wu C, Li X, Song Y, Yao X, Wu X, Duan Y, Zhang H, Wang Y,
QianZ,CuiJ,LuJ. 2020.Ontheorigin andcontinuingevolution ofSARS-
CoV-2.National Science Review,doi:10.1093/nsr/nwaa1036.
The Asahi Shimbun. 2020 (2020-02-02). Japan tightens immigration as 3
more infected by coronavirus. http://www.asahi.com/ajw/articles/AJ2020
02020013.html.
Wikipedia. 2020 (2020-02-29). 2019-20 coronavirus outbreak. https://en.
wikipedia.org/wiki/2019%E2%80%9320_coronavirus_outbreak.
WongG, BiYH,WangQH, ChenXW,Zhang ZG,YaoYG. 2020.Zoonotic
origins of human coronavirus 2019 (HCoV-19 / SARS-CoV-2): why is this
workimportant?.Zoological Research,41(3):213−219.
WuF, ZhaoS,Yu B,Chen YM,Wang W,Song ZG,Hu Y,Tao ZW,Tian
JH,PeiYY,YuanML,ZhangYL,DaiFH,LiuY,WangQM,ZhengJJ,XuL,
HolmesEC, ZhangYZ.2020a. Anew coronavirusassociated withhuman
respiratorydiseaseinChina.Nature,579(7798):265−269.
Wu JT, Leung K, Leung GM. 2020b. Nowcasting and forecasting the
potential domestic and international spread of the 2019-nCoV outbreak
originating in Wuhan, China: a modelling study. The Lancet, 395(10225):
689−697.
WuZY,McGooganJM.2020.Characteristicsofandimportantlessonsfrom
thecoronavirusdisease2019(COVID-19)outbreakinChina:summaryofa
report of 72 314 cases from the Chinese Center for Disease Control and
Prevention.JAMA,doi:10.1001/jama.2020.2648.
XiaoKP,ZhaiJQ,FengYY,ZhouN,ZhangX,ZouJJ,LiN,GuoYQ,LiXB,
ShenXJ,ZhangZP,ShuFF,HuangWY,LiY,ZhangZD,ChenRA,WuYJ,
PengSM, HuangM, XieWJ,Cai QH,Hou FH,LiuYH, ChenW, XiaoLH,
Shen YY. 2020. Isolation and characterization of 2019-nCoV-like
coronavirus from malayan pangolins. bioRxiv, doi: 10.1101/2020.02.17.
951335.
YangY, LuQB, LiuMJ,Wang YX,Zhang AR,JalaliN, DeanN, LonginiI,
Halloran ME, Xu B, Zhang XA, Wang LP, Liu W, Fang LQ. 2020.
Epidemiological and clinical features of the 2019 novel coronavirus
outbreakinChina.medRxiv,doi:10.1101/2020.02.10.20021675.
Yao YG, Kong QP, Bandelt HJ, Kivisild T, Zhang YP. 2002.
Phylogeographic differentiation of mitochondrial DNA in Han Chinese.
American Journal of Human Genetics,70(3):635−651.
Zhang RQ, Liu H, Li FY, Zhang B, Liu QL, Li XW, Luo LM. 2020a.
Transmission and epidemiological characteristics of Novel Coronavirus
(2019-nCoV) Pneumonia (NCP): preliminary evidence obtained in
comparisonwith2003-SARS.medRxiv,doi:10.1101/2020.01.30.20019836.
Zhang T, Wu QF, Zhang ZG. 2020b. Pangolin homology associated with
2019-nCoV.Current Biology,30:1346−1351.
ZhaoZM, LiHP, WuXZ, ZhongYX, ZhangKQ, ZhangYP, BoerwinkleE,
FuYX.2004.ModeratemutationrateintheSARScoronavirusgenomeand
itsimplications.BMC Evolutionary Biology,4:21.
ZhouP,YangXL, WangXG,HuB,ZhangL, ZhangW,SiHR,ZhuY,LiB,
HuangCL, ChenHD, ChenJ, LuoY, GuoH, JiangRD, LiuMQ, ChenY,
ShenXR, WangX,Zheng XS,Zhao K,Chen QJ,Deng F,Liu LL,Yan B,
Zhan FX, Wang YY, Xiao GF, Shi ZL. 2020. A pneumonia outbreak
associated with a new coronavirus of probable bat origin. Nature,
579(7798):270−273.
ZhuN,ZhangDY,WangWL, LiXW,YangB,SongJD,ZhaoX,HuangBY,
ShiWF, LuRJ,Niu PH,ZhanFX,MaXJ,WangDY, XuWB,Wu GZ,Gao
GF, Tan WJ. 2020. A novel coronavirus from patients with pneumonia in
China,2019.The New England Journal of Medicine,382(8):727−733.
Zoological Research41(3):247−257,2020 257