Mitochondrial and ribosomal markers in the identification of nematodes of clinical and veterinary importance

Background Nematodes of the Ascarididae, Ancylostomatidae and Onchocercidae families are parasites of human and veterinary importance causing infections with high prevalence worldwide. Molecular tools have significantly improved the diagnosis of these helminthiases, but the selection of genetic markers for PCR or metabarcoding purposes is often challenging because of the resolution these may show. Methods Nuclear 18S rRNA, internal transcribed spacers 1 (ITS-1) and 2 (ITS-2), mitochondrial gene cytochrome oxidase 1 (cox1) and mitochondrial rRNA genes 12S and 16S loci were studied for 30 species of the mentioned families. Accordingly, their phylogenetic interspecies resolution, pairwise nucleotide p-distances and sequence availability in GenBank were analyzed. Results The 18S rRNA showed the least interspecies resolution since separate species of the Ascaris, Mansonella, Toxocara or Ancylostoma genus were intermixed in phylogenetic trees as opposed to the ITS-1, ITS-2, cox1, 12S and 16S loci. Moreover, pairwise nucleotide p-distances were significantly different in the 18S compared to the other loci, with an average of 99.1 ± 0.1%, 99.8 ± 0.1% and 98.8 ± 0.9% for the Ascarididae, Ancylostomatidae and Onchocercidae families, respectively. However, ITS-1 and ITS-2 average pairwise nucleotide p-distances in the three families ranged from 72.7% to 87.3%, and the cox1, 12S and 16S ranged from 86.4% to 90.4%. Additionally, 2491 cox1 sequences were retrieved from the 30 analyzed species in GenBank, whereas 212, 1082, 994, 428 and 143 sequences could be obtained from the 18S, ITS-1, ITS-2, 12S and 16S markers, respectively. Conclusions The use of the cox1 gene is recommended because of the high interspecies resolution and the large number of sequences available in databases. Importantly, confirmation of the identity of an unknown specimen should always be complemented with the careful morphological examination of worms and the analysis of other markers used for specific parasitic groups. Graphical Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s13071-023-06113-4.


Background
Nematodes of the families Ascarididae, Ancylostomatidae and Onchocercidae are parasites of human and veterinary importance causing infections with high prevalence worldwide [1][2][3].Many of the infections produced in humans are cataloged as tropical neglected diseases, associated to areas with low socioeconomic index and poor access to diagnostic tools and treatment [4,5].
Members of the families Ascarididae, Ancylostomatidae and Onchocercidae may cause clinical manifestations in virtually all known mammal species, whom get infected by the intake of contaminated water or food, contact with soil or an intermediate host [6].For instance, the human ascarid Ascaris lumbricoides is the most common human nematode reported worldwide, with up to 1.2 billion estimated people infected [7].In the veterinary field, Parascaris equorum and Baylisascaris procyonis produce intestinal infections in equines and raccoons [8], respectively, whereas Toxocara canis generates infections in > 100 million dogs worldwide [9] and can produce a zoonotic disease in humans, known as visceral, ocular or cerebral larva migrans with an estimated global human seroprevalence of 19% [9,10].Furthermore, blood-sucking hookworms like Necator americanus and Ancylostoma spp.(Ancylostomatidae family) are responsible for 740 million people infected worldwide causing intestinal bleeding and chronic ferropenic anemia [4].Finally, filarioids such as Wuchereria bancrofti, Brugia malayi, Onchocerca spp.and Dirofilaria spp.are vector-borne heteroxenous parasites of the family Onchocercidae with worldwide distribution that affect thousands of humans, companion animals, livestock and wildlife [5].
Sensitive diagnostic tools are essential to improve clinical detection and treatment in animal and human hosts.However, traditional parasitological methods of diagnosis usually show low sensitivity for detecting and identifying species [3].For instance, although direct microscopic observation of stool samples and concentration methods like Kato-Katz, Sheather or Ritchie detect ascarid and hookworm eggs and larvae using unsophisticated equipment and inexpensive reagents [11], these methods are negatively affected by low parasitic loads, intermittent excretion of eggs and uneven distribution of eggs in stool samples, significantly affecting their sensitivity [12].Similarly, the diagnosis of filariasis is based predominantly on the microscopic detection of circulating microfilariae in blood, either by direct blood smears or Knott's modified test, or the recognition of the worm antigens using immunochromatographic tests.However, false-negative results are common in low parasitemia, single-sex or mixed infections.To alleviate these limitations of classical methods in parasitology, molecular methods with higher sensitivity and specificity have been designed [13,14].
Molecular tools have improved the diagnosis of infections and facilitated taxa differentiation and phylogenetic analysis of worms of the Ascarididae, Ancylostomatidae and Onchocercidae families [14,15].Mitochondrial genes such as cytochrome c oxidase subunit 1 (cox1), cytochrome b (cytb), and the NADH dehydrogenase subunit 1 (nad1), and mitochondrial rRNA 12S and 16S are widely used to determine lower taxonomic levels (i.e.identification to genus and species-level) and resolve species-level phylogenies because of their high degree of sequence variation [16].Similarly, internal transcribed spacers (ITS) 1 and 2 of the ribosomal nuclear DNA also show a high degree of sequence variation, promoting their use as markers for species differentiation [16].Contrarily, the 18S nuclear rRNA gene is highly conserved within species and has low inter-species variability but has been employed for characterizing nematode communities at higher taxonomic levels [16].Altogether, the variability of each genetic marker offers different resolution for phylogenetic analyses and practicability for diagnosis of nematodes [15].Some challenges remain unsolved in the genetic analysis of nematodes.For instance, species hybrids have been initially interpreted as sequencing artifacts when discordances are detected during phylogenetic analyses using nuclear and mitochondrial genes [17].Moreover, cryptic diversity identification in morphologically undistinguishable specimens relies on nucleotide differences among taxa and sequence availability in databases [18].Therefore, cryptic diversity is accurately detected only when enough sequences are compared and when appropriate nuclear, ribosomal or mitochondrial markers are employed.
The aim of this study was to analyze the resolution and performance of six different ribosomal and mitochondrial loci (18S, ITS-1, ITS-2, cox1, 12S and 16S) for the interspecies differentiation of nematodes of the families Ascarididae, Ancylostomatidae and Onchocercidae of human and veterinary importance.Importantly, we estimated the number of available sequences on databases for each species included in the study to contextualize their usefulness in the clinical and research scenario.In this way, we evaluated the practicality of ribosomal and mitochondrial markers for the identification of worms as well as their ease of analysis and taxonomic resolution.

Methods
Six genetic markers from representative nematode species of medical and veterinary importance of the families Ascarididae, Onchocercidae and Ancylostomatidae were analyzed in this study (Fig. 1).Accordingly, five representative full-length or partial sequences of > 300 bp nuclear ribosomal regions ITS-1, ITS-2 and 18S rRNA and mitochondrial genes cox1, 12S and 16S rRNA were downloaded from the NCBI database (https:// www.ncbi.nlm.nih.gov/) and aligned in MEGA X using the MUSCLE tool.Chosen sequences had to overlap with the consensus sequence block in at least 80% of length of the block to be included in the analysis.To take into account intraspecies variability, up to five sequences belonging to different hosts or geographical locations were included in the analysis.The accession numbers of the sequences included in this study are available in Additional file 1: Table S1.The best nucleotide substitution model for each loci and family was computed with jModelTest software [19] based on the Bayesian inference criterion calculations (Additional file 2: Table S2).Bayesian inference (BI) phylogenetic trees were constructed using the BEAST2 package (https:// www.beast2.org/) with 10 8 MCMCs and a tree burn-in percentage of 10%.Then, the program Interactive Tree of Life (iTOL) (https:// itol.embl.de/) was used to edit the resulting trees.Afterwards, we elaborated a color-coded matrix based on sequence pairwise identity using the program Sequence Demarcation Tool version 1.2 (SDTv1.2) [20], with alignments as input files and removing the option for calculating a neighbor-joining tree.
Violin plots of genetic p-distances were constructed using the R package ggplot2 [21].Pairwise nucleotide pdistances were calculated in MEGA X.Data and script used for this purpose are available in the Additional file 3. The Gaussian distribution of data was verified using the Kolmogorov-Smirnov test to determine significance between pairwise p-distance values between markers.Then, a Kruskal-Wallis analysis for non-parametric data was performed followed by Dunn's multiple comparison test, and significance was considered when P < 0.01.These analyses were run in GraphPad Prism version 10.

Sequence availability
The total number of sequences from the abovementioned species and genetic markers were calculated using the NCBI database and BLAST tool (https:// blast.ncbi.nlm.nih.gov/ Blast.cgi).The search was done to assess the practicality of each marker in the context of a diagnostic challenge.BLAST search was done with a complete sequence for each species.If there was no reference sequence for a particular locus, the longest sequence was selected.To retrieve all sequences available in the NCBI database, all sequences obtained from the blastn search of the reference sequences were downloaded.In addition, a manual search using key words for each genetic marker and species was also performed to retrieve more divergent sequences.Afterwards, we proofread all sequences and maintained only those with a length > 300 bp and > 97% percentage of identity with the reference sequence.

Analysis of 18S rRNA
The analyses for the 18S rRNA region showed a minimum paired identity > 96.3% for the three studied families (Fig. 2).Accordingly, the lowest nucleotide similarity within Ascarididae species was 98.3%, in Ancylostomatidae was 99.60% and for Onchocercidae members was 96.3%.
The BI analysis revealed low posterior probabilities (PP) in the branches separating several species of the Ascarididae and Ancylostomatidae families, indicating sequences could not be resolved and separated with enough resolution.Therefore, the phylogenetic analysis with the 18S does not present a clear differentiation of the species of these families (Fig. 2a and b).When Toxocara canis, T. cati and T. vitulorum were compared, nucleotide similarities were > 99%, and all sequences retrieved from these species grouped in the same cluster with low PP in the inner ramifications.For the Ancylostomatidae family, only ten sequences could be obtained from GenBank, and these were divided in a tree branch containing the sequences of Ancylostoma spp.and N. americanus, while another tree branch contained Uncinaria stenocephala specimens (Fig. 2b), with low PP in inner ramifications.For the Onchocercidae, four main clusters were observed in the BI tree: one with Dirofilaria repens, Loa loa and Onchocerca fasciata; a second with W. bancrofti, B. malayi and a D. repens sequence from a dog from France; a third group with Mansonella spp.; a final cluster with Dirofilaria immitis sequences (Fig. 2c).

Analysis of ITS-1 and ITS-2 ribosomal loci
The ITS regions showed high interspecies resolution with nucleotide pairwise identities > 51.2%.The ITS-1 region had an average minimum pairwise identity of 51.2%, 57.9% and 58.3% for the Ascarididae, Ancylostomatidae and Onchocercidae families, respectively, with complete separation of species in the BI trees (Additional file 4).Moreover, the ITS-2 showed 52.5% mean nucleotide pairwise similarity in the Ascarididae family, 72.8% for Ancylostomatidae and 51.5% in the Onchocercidae family (Fig. 3).
A total separation of species of the Ascarididae was observed in the ITS-1 and ITS-2 BI tree families, supported by PP values of 1.000 (Additional file 4 and Fig. 3).The ITS-1 and ITS-2 BI trees showed two clear clusters: one containing Toxascaris leonina, Ascaris lumbricoides, A. suum, P. equorum, Baylisascaris procyonis and B. transfuga sequences and a second containing T. canis, T. cati and T. vitulorum sequences.Remarkably, A. lumbricoides sequences from humans and A. suum, obtained from pigs and a human, clustered together with a minimum identity percentage of 98.4% for the ITS-1 and 100.0%for ITS-2.In addition, sequences of these two ascarids were intermixed in the cluster in the ITS-2 BI tree (Fig. 3a).
Four main groups could be identified in the ITS-1 and ITS-2 phylogenetic trees of the Ancylostomatidae family: one represented by Ancylostoma spp., a second represented by Uncinaria sanguinis, one represented by Bunostonum spp.and a group represented by Necator americanus sequences, all with PP = 1.000.Moreover, each species within each cluster was separated between them, showing a clear interspecies resolution.
For the Onchocercidae family, four sequence groups were observed in the BI tree of the ITS-1 region.Accordingly, one group contained Mansonella spp.sequences, the second contained Onchocerca spp., the third contained W. bancrofti and B. malayi and, lastly, a cluster contained Dirofilaria spp.and L. loa sequences (Additional file 4).Each species was separated with high nucleotide difference, except for W. bancrofti and B. malayi, whose differences were approximately 10%.ITS-2 phylogenetic tree built with sequences of this family exhibited the same clustering as ITS-1, with the addition of Acanthocheilonema reconditum sequences, which were located next to the Mansonella spp. group.Although four ITS-1 sequences of A. reconditum were available in GenBank, these did not meet the criteria to be included in the BI analysis.

cox1 mitochondrial gene
Average pairwise nucleotide similarities in the cox1 gene fragment within the Ascarididae, Ancylostomatidae and Onchocercidae families were 85.4%, 83.0% and 80.5%, respectively (Fig. 4).All species used in the analysis were successfully resolved in the phylogenetic tree for ascarids and showed a similar topology with the BI phylogenetic tree using the ITS-1 and ITS-2.Accordingly, three main clusters were observed: one with A. lumbricoides, A. suum, T. leonina and Lagochilascaris minor, a second group with B. transfuga, B. procyonis and P. equorum and a final one with Toxocara spp.Finally, as previously observed in the ITS-1 and ITS-2 trees, A. lumbricoides and A. suum clustered together, resulting in very high paired identity values, with the lowest value being 96.9% and the highest 100.0%(Fig. 4a).Moreover, two groups were observed in the Ancylostomatidae family tree, namely one with Ancylostoma spp.and U. stenocephala and another cluster with N. americanus and Bunostomum spp.(Fig. 4b).Phylogenetic analysis of the Onchocercidae sequences showed a marked divergence from Mansonella spp.regarding the other species of the family with high PP.In addition, Acanthocheilonema reconditum sequences were grouped with those of Onchocerca spp., in contrast to the topologies for the ITS-1 and ITS-2 trees, which were separated into different groups (Fig. 4c).

12S and 16S mitochondrial genes
Phylogenetic interspecies resolution was obtained in the analyzed 12S and 16S fragments (Additional files 5 and 6, respectively).The average pairwise nucleotide similarity in the 12S was 75.9% within the Ascarididae family, 84.7% for Ancylostomatidae and 81.5% within the Onchocercidae.On the other hand, the average pairwise nucleotide similarity for the 16S gene was 79.0%, 83.1% and 83.5% within the analyzed ascarid, ancylostomatid and onchocercid sequences.In addition, the number of sequences available for phylogenetic analysis for the 16S gene was lower than in the other loci.
Phylogenetic trees for 12S and 16S genes support the results previously obtained for A. lumbricoides and A. suum, which grouped together with a percentage of identity for the 16S sequences ≥ 97.0%.Furthermore, A. lumbricoides, A. suum, P. equorum, B. procyonis, B. transfuga and T. leonina were grouped in one of the branches of the tree, while Toxocara spp.sequences were clustered in a separate cluster with high PP and pairwise identities ranging from 77.0 to 79.0% compared to sequences of other ascarids.
In the 12S phylogenetic tree, two main branches could be observed, one with N. americanus sequences and the second with Bunostonum spp., Ancylostoma.spp.and Uncinaria spp.The tree topology for 16S sequences varied in that Bunostomum spp.sequences were separated from the other analyzed species, whereas N. americanus sequences were closer to Ancylostoma spp.
The phylogenetic trees of the 12S and 16S of the Onchocercidae sequences revealed a different topology to the one observed in the cox1 gene fragment.For the 12S tree, three groups were observed: a branch with Onchocerca spp.and Dirofilaria spp., a second group with B. malayi, W. bancrofti, L. loa and Mansonella spp., and a third cluster with A. reconditum.In the 16S phylogenetic tree, fewer sequences could be analyzed according to the inclusion criteria.Nevertheless, a group with B. malayi, W. bancrofti, L. loa and O. volvulus was obtained; a separate cluster was observed with Dirofilaria spp.and a single M. perstans sequence was derived from the tree.Therefore, D. immitis and D. repens were grouped together in both 12S and 16S genes, and the same was obtained with B. malayi and W. bancrofti.

Nucleotide differences between genes of each family
Significant pairwise nucleotide p-distances in all markers were obtained for each family (all P < 0.0001, Kruskal-Wallis statistic = 2023, 208 and 1900 for Ascarididae, Ancylostomatidae and Onchocercidae families, respectively) (Fig. 5).A total of 4998 pairwise comparisons were calculated for the Ascarididae family, 2828 for the Ancylostomatidae family and 4038 for the Onchocercidae family.
The highest pairwise nucleotide similarity in the Ascarididae family was shown in the nuclear rRNA 18S fragment with an average of 99.1 ± 0.6%, which was significantly different from the other five markers (all P < 0.01) (Fig. 5a).Average pairwise nucleotide p-distances of the analyzed cox1 fragments were 90.4 ± 3.1%, whereas the pairwise nucleotide similarities in the 16S, 12S, ITS-1 and ITS-2 markers were 87.0 ± 6.4%, 86.4 ± 6.3%, 85.0 ± 7.9% and 72.9 ± 14.1%, respectively.Significant differences were detected when comparing the values of each marker between them, except between average pairwise p-distances of 12S and 16S, 16S and ITS-1, and 12S and ITS-1.
For the Ancylostomatidae family, the highest pairwise similarity was obtained in the 18S fragment, with an average of 99.8 ± 0.1% similarity between sequences, followed by the 12S with 90.3 ± 3.9%, the 16S with 88.7 ± 4.5%, cox1 with 88.7 ± 3.6%, ITS-1 with 87.3 ± 9.4% and ITS-2 with 86.4 ± 7.9% average pairwise similarities (Fig. 5b).The average pairwise p-distance was significantly different when comparing the data of the 18S to the other five markers with all P < 0.01.Other pairwise statistical comparisons were significant except between data of 16S and cox1, cox1 and ITS-1, and ITS-1 and ITS-2, whose average pairwise distance where similar between them.

Availability of sequences in the GenBank database
The marker with the highest representation in GenBank was the cox1 with a total of 2491 sequences for the evaluated species, followed by the ITS-1 (n = 1082), ITS-2 (n = 994), 12S (n = 428), 18S (n = 212) and 16S (n = 143) (Fig. 6).Complete results with the number of sequences per analyzed species is available in Additional file 7. Accordingly, the analyzed species of the Ascarididae family had 870 cox1 sequences deposited in GenBank, whereas 525 sequences were available for the species of the Ancylostomatidae family and 1096 for the Onchocercidae family.For the ITS-1, 498, 375 and 209 sequences were retrieved from GenBank in the studied species of the Ascarididae, Ancylostomatidae and Onchocercidae families, respectively.The marker with the least sequence availability was the 16S mitochondrial rRNA gene with 46 sequences in the analyzed species of the Ascarididae family, 56 for the Ancylostomatidae and 41 for the Onchocercidae.The numbers of sequences per species are presented in Table 1.

Discussion
We analyzed the interspecies resolution and sequence availability of different mitochondrial and ribosomal markers commonly used for diagnosing helminthiases of medico-veterinary significance.The present work determined that the mitochondrial gene cox1 had the highest number of available sequences in GenBank, indicating this is a suitable marker to use when confirming the identity of an unknown specimen since many sequences are available for comparison.Moreover, this analysis included sequences from different hosts and geographical regions, thus giving a wider perspective of the genetic variability of each species.The nuclear rRNA 18S gene showed the lowest species resolution, whereas nuclear ITS-1 and ITS-2 regions showed the highest nucleotide differences between species, followed by mitochondrial gene cox1 and mitochondrial rRNA genes 12S and 16S.Furthermore, the cox1 gene was the most represented The results obtained for the 18S gene of the Ascarididae, Ancylostomatidae and Onchocercidae families indicate that this marker does not separate between close species, since nucleotide similarity values between 98.8 and 99.9% were obtained.The analyzed species were intermixed in the phylogenetic trees of the three families, especially between A. lumbricoides and A. suum, Toxocara spp. with T. leonina, W. bancrofti and B. malayi, and among Mansonella spp.Therefore, this marker may resolve species at higher taxonomic levels [15,22,23], which may be useful for describing nematode communities in soil [24,25] or biological animal samples [26] using 18S-based metagenomics.However, it does not provide the necessary resolution for identifying unknown specimens in a clinical setting [27].For instance, an unknown specimen may be erroneously classified if only an 18S fragment is analyzed, such as the case of L. minor from cats, which showed 99% homology to A. suum [27].For these cases, the analysis of several and longer DNA fragments is recommended, together with the close morphological examination of specimens.The low interspecies difference in the 18S may be explained by the concerted evolution of this coding region, which homogenizes all copies in a specimen's genome [28] and secures the structure of rRNA functional domains, as opposed to the non-coding ITS regions [29].
The ITS-1 and ITS-2 regions yielded the highest percentage of sequence variability and separated all species in phylogenetic analyses except A. suum and A. lumbricoides, which were intermixed and Ancylostoma spp. that showed lower nucleotide differences between them.Although the ITS regions are part of the nuclear rRNA operon, these loci have a higher nucleotide substitution rate than the 18S [29], which eases differentiation at the species level [15].This characteristic was confirmed in the present study and has been observed for nematodes of other families like Spirocercidae [30], Strongylidae [31] and Trichuridae [32], to name some examples.Nucleotide differences in these loci are so high that they may even exhibit a high genetic intra-individual variation, i.e. a single specimen with different copies of ITS, such as the case of the dog carcinogenic nematode Spirocerca lupi [30], the whipworms Trichuris spp.[33] or the causing agent of cystic echinococcosis, Echinococcus granulosus [34].In this study, ITS-1 and ITS-2 proved to be useful markers in helminth identification applications.Nevertheless, the high nucleotide difference between species difficulted subsequent analyses, such as sequence alignment and phylogenetic reconstructions when using large datasets [15].For this reason, ideal diagnostic markers would be those showing enough interspecies nucleotide differences but easy to handle for beginners.
The three markers of mitochondrial origin analyzed in this work had a similar performance in terms of their sequence resolution in phylogenetic analyses and the pairwise nucleotide differences.It has been argued that mitochondrial rRNA 12S and 16S are less variable than cox1.However, we did not observe significant statistical differences in the percentage of nucleotide variation in the three markers [15].The cox1 gene has been used in DNA barcoding and is currently considered the universal genetic marker for this application [35].DNA barcoding uses a small consensus DNA fragment of the cox1 to globally identify a large number of organisms from different geographical locations, hosts or genetic subclassifications [35][36][37].Mitochondrial markers have several advantages compared to those of nuclear origin.For example, mitochondrial markers do not undergo recombination and have a higher mutation rate than nuclear rRNA 18S, 5.8S or 28S [38].Therefore, mitochondrial genes are useful in resolving organisms at lower taxonomic levels to the species level and thus may be useful for metabarcoding purposes.
Although mitochondrial markers have great variability, it has been argued that the design of universal primers targeting mitochondrial regions may be challenging [39].However, previous studies have successfully amplified consensus cox1 sequences from different nematode families [40] and have described free-living marine nematode communities [41].Moreover, a study of seven nematode genera across different families showed an interspecies mean p-distance or nucleotide difference in the cox1 between 10 and 25%, whereas intraspecific pairwise nucleotide differences ranged from 0.5 to 3.5% [39].Those results resemble the ones obtained in the present study.Altoghether, our results support the applicability of cox1 for metabarcoding studies.
The number of cox1 sequences available in GenBank exceeded those of the other five markers according to the employed mining pipeline.The present study analyzed this aspect for the first time, thus giving a different perspective regarding the usefulness of each locus in a diagnostic context.The larger number of cox1 sequences may be explained by the high sensitivity of the primers used for amplifying this gene, the widespread use of certain primers targeting the cox1 and the efforts of research groups worldwide in using and depositing barcodes of their specimens, thus including all possible biodiversity [39].Interspecies resolution obtained in the 12S and 16S rRNA mitochondrial genes were comparable to those of the cox1 and therefore may have similar utility in metabarcoding analyses.However, the use of 12S and 16S and their applicability for phylogenetic, taxonomic and barcoding analysis would be limited by the low number of sequences in databases.A similar situation was observed with ITS regions, which provide larger nucleotide differences between species compared to nucleotide markers as obtained before [15], but their lower availability would limit their use for diagnostic or metabarcoding purposes.Therefore, sequencing and deposition of cox1, 12S, 16S, ITS-1 and ITS-2 sequences from nematode specimens are strongly encouraged to include the most comprehensive genetic information, which will be useful in future studies.
Interestingly, A. lumbricoides and A. suum sequences were intermixed in all analyzed markers, thus raising the long-standing question of whether these two taxonomic units actually represent the same species [42].Cryptic diversity is defined as two morphologically indistinguishable but genetically different entities [43].Additionally, other parameters have been used to infer a cryptic species including host preference, geographic distribution, resource use and reproductive behavior [18].Accordingly, both ascarids have been proposed as cryptic species since A. suum was named as a pig roundworm by Goeze (1782), even though it did not present any morphological differences from A. lumbricoides.The present study showed that A. suum and A. lumbricoides clustered together in the phylogenetic trees of the different genes and had a pairwise nucleotide identity between 97 and 100%.In fact, genomic analyses of A. lumbricoides obtained from humans of a village without pig husbandry showed a high heterozygocity and hybridization of A. lumbricoides and A. suum, suggesting a potential interbred Ascaris complex infecting humans [44].In any case, an allele-specific PCR using the ITS-1 of A. suum and A. lumbricoides showed good resolution for discriminating both ascarids, which may become useful for epidemiological and control purposes [45].
The selection of markers used in the present study was not exhaustive, and they were chosen according to their current use for diagnostic purposes.Other mitochondrial genes such as the nad1 and cytb have been evaluated before [15], rendering similar interspecies variation as obtained herein for cox1, 12S and 16S analyses.Additionally, representative nuclear and mitochondrial sequences of nematodes, trematodes and cestodes of clinic and veterinary importance have been evaluated before.Nevertheless, we focused on 30 species of three nematode families of medico-veterinary importance, rendering up to five sequences from each species when available.In this way, genetic variation arising from geographic or host origin could be included herein, which has not been studied before to our knowledge.Importantly, these kind of studies are necessary to guide basic and medical researchers in their decision-making when unknown worms arrive at their laboratories or clinics.
Even though molecular diagnosis of nematodes may be objective and straightforward [46], a proper nematode identification should include morphological analysis of specimens [14].Molecular data complement morphological observations of worms, which may lead to the identification of cryptic diversity sensu lato or sensu stricto, and the reclassification of certain taxa or the description of new species [18].This integrative taxonomic approach has been suggested before for filarioids [14] and has proven essential for the description of the true biodiversity of a genus, family or phyla.Furthermore, incorrect deposition of genetic data has been demonstrated in other invertebrate orders [47] and for the dyphyllobothridean cestodes of the genus Spirometra [48].This highlights the responsibility of researchers in the selection of methods and the interpretation of their results.

Conclusions
Phylogenetic relationships and minimal pairwise nucleotide identities were analyzed in six different molecular markers from 30 species of the Ascarididae, Ancylostomatidae and Onchocercidae families.Importantly, we assessed, for the first time to our knowledge, the availability of those sequences in databases as a complement for choosing a molecular marker in a clinical or research context.Results showed that cox1 is a useful marker for diagnostic or metabarcoding purposes due to the high interspecies resolution demonstrated in phylogenetic analyses and the higher number of sequences available in databases.This will enable the accurate identification of an unknown specimen when comparing their associated sequences to those deposited in GenBank.Finally, we recommend the combined analysis of ITS regions, other mitochondrial genes and morphological analysis of specimens whenever possible.

Fig. 1
Fig.1Ribosomal and mitochondrial sequences used for the phylogenetic analysis of 30 species of the Ascarididae, Ancylostomatidae and Onchocercidae families.The number of sequences is color coded from darkest (n = 5) to lightest (n = 0) according to the marker: 18S in blue, ITS1 in purple, ITS2 in red, cox1 in yellow, 12S in green and 16S in turquoise.This figure was created using Biorender.com

Fig. 2
Fig. 2 Phylogenetic Bayesian inference analysis of the nuclear 18S rRNA of selected species of veterinary and clinic importance of the families Ascarididae (a), Ancylostomatidae (b) and Onchocercidae (c).Nucleotide pairwise identity is denoted in the heatmap next to each taxa name

Fig. 5 Fig. 6
Fig. 5 Nucleotide similarities between selected species of veterinary and clinic importance of the families Ascarididae (a), Ancylostomatidae (b) and Onchocercidae (c) in each marker analyzed in this study

Table 1
Number of sequences retrieved from GenBank of nematodes of veterinary and clinical importance of the Ascarididae, Ancylostomatidae and Onchocercidae families