Reply to “May Newly Defined Genotypes XVII and XVIII of Newcastle Disease Virus in Poultry from West and Central Africa Be Considered a Single Genotype (XVII)?”

REPLY

As members of the Working Group on Newcastle disease virus (NDV) nomenclature, we welcome any contribution to the discussion and definition of genotypes. We have recently published data on new NDV strains from West and Central Africa, some of which were assigned to new genotypes. Our analysis showed that genotypes XVII and XVIII fulfill all criteria for the definition of new genotypes (Table 1), including their mean evolutionary distance of >0.10 (Tables 1 and 2). These conclusions are based on the recommendations of Diel et al. (1).

TABLE 1.

Conformity of the grouping of genotypes XVII and XVIII into two distinct genotypes in compliance with the criteria defined by Diel et al. (1)

Criterion defined by Diel et al. (1)	Conformity of genotypes XVII and XVIII definition according to the criteria
New genotype grouping supported by the tree topology	Genotypes XVII and XVIII form 2 distinct groups of strains as shown on the phylogenetic tree (see the supplemental material provided with the original manuscript [3])
New (sub)genotype(s) designated only when complete F gene sequences from at least four independent isolates are available	Genotype XVII includes 56 strains found in Central African Republic in 2008, in Nigeria from 2007 to 2011, in Niger in 2008, in Benin in 2009, in Burkina Faso in 2008, and in Cameroon in 2008; genotype XVIII includes 15 strains found in Mali in 2007–2009, in Mauritania in 2006, in Ivory Coast in 2006–2008, in Togo in 2009, and in Nigeria in 2011
Mean evolutionary distance between genotypes > 0.10	0.111 (see Table 2)
Bootstrap value at the genotype defining node > 60%	Bootstrap value at the node defining genotype XVII = 99%; bootstrap value at the node defining genotype XVIII = 99%

TABLE 2.

Estimates of net mean evolutionary distance and mean evolutionary distance between NDV genotypes^a

^aThe “net mean evolutionary distances” between groups of sequences are shown in white, and the “mean evolutionary distances” are shown in gray above the diagonal. Values below the 0.10 threshold are shown in bold characters. As calculated in MEGA6, the “net mean evolutionary distance” between genotypes X and Y is equal to d_XY − [(d_X + d_Y)/2], where d_XY is the “mean evolutionary distance” between genotypes X and Y and d_X and d_Y are the mean within-genotype distances. Analyses were conducted using the maximum composite likelihood model. The rate variation among sites was modeled with a gamma distribution (shape parameter = 1). The analysis involved 873 nucleotide sequences. Codon positions included were 1st plus 2nd plus 3rd plus noncoding. All ambiguous positions were removed for each sequence pair. There were a total of 1,662 positions in the final data set. Evolutionary analyses were conducted in MEGA6.

In their letter, Desingu et al. challenge our conclusion and suggest that genotypes XVII and XVIII should be grouped in a single genotype (2). Their conclusion was based on “net mean evolutionary distances” (0.071, as defined in MEGA6; see definition in legend of Table 2) below the 0.10 threshold. However, current nomenclature recommendations and the 0.10 threshold are based on “mean evolutionary distances.” Using our data set, the mean evolutionary distance between genotypes XVII and XVIII is above the 0.10 threshold for separate genotypes, regardless of the distribution of rate variations among sites used (0.101 with uniform rates among sites; see Table 3 of our article [3]; 0.111 with rate variation among sites modeled with a gamma distribution and shape parameter = 1; Table 2). Thus, matching the correct parameter with the correct threshold shows that genotypes XVII and XVIII are distinct.

We show below that there is a significant difference between the “net mean evolutionary distance” (Table 2, below the main diagonal) and the “mean evolutionary distance” (Table 2, above the main diagonal) for every combinatorial comparison of genotypes: 34 “net mean evolutionary distances” have a value below the 0.10 threshold (Table 2, values in bold), while all but 1 of the “mean evolutionary distances” are >0.10. These data show that using the “net mean evolutionary distances” as suggested by Desingu et al. would not make sense since a large number of genotypes would be merged, although some of these do not fall into monophyletic clusters (see Fig. S1 in the supplemental material provided with the original article [3]).

Furthermore, Desingu et al. did not even consider updating the number of strains included in the analyses. Indeed, since our manuscript was accepted for publication in the Journal of Clinical Microbiology, a large number of NDV sequences have been submitted to GenBank. Over 1,600 sequences of good quality can now be used for classification. Among others, four new strains from Nigeria are now classified into genotype XVII (4) and two new strains from Mali have been assigned to genotype XVIII.

In brief, Desingu et al. apply the wrong parameter with the right threshold. The analysis included here confirms that genotypes XVII and XVIII have to be considered distinct genotypes.