SSR and SNP genetic profiling of Armenian grape cultivars gives insights into their identity and pedigree relationships

The South Caucasus is recognised as the primary Vitis vinifera L. (grapevine) domestication centre and has a high diversity of wild and cultivated grapevines. Archaeological findings indicate that winemaking activities have existed in Armenia for more than 6,000 years, viticulture being one of the most important activities of the modern Armenian agricultural sector. Despite this relevance, some grapevines in local collections have not yet been properly identified, thus hindering the efficient conservation, characterisation and eventual use of autochthonous genetic resources. In the present study, a combined SNP and SSR profiling strategy was used for the genetic identification of a series of grapevine accessions from the Grape Collection of the International Academy of Viticulture and Winemaking in Nalbandyan, presumed to be autochthonous Armenian varieties. The results provided useful information for the correct identification of these genetic resources, revealing multiple cases of synonyms, homonyms and misnames. The genetic data made it possible to confirm the pedigree proposed for some of the cultivars identified in this study and to clarify the origin of others. In addition, we propose, for the first time, a series of new trios and duos involving autochthonous Armenian grapevines. The singularity of this genetic pool compared to other Western and Central European varieties, as well as the potential novel sources of variability in traits of interest (e.g., seedlessness) that were found, highlight the importance of improving knowledge of the Armenian grapevine genetic pool.


INTRODUCTION
The South Caucasus is acknowledged as the primary grapevine (Vitis vinifera L.) domestication center . The genetic characterisation of grapevines from this region indicates the uniqueness and singularity of this germplasm when compared to Western genetic pools Imazio et al., 2013;Maul et al., 2015;Riaz et al., 2018), representing an important source of diversity for future breeding programmes and sustainable agriculture (Dallakyan et al., 2020). Analyses of South Caucasus grapevines by molecular markers show a moderate genetic differentiation between V. vinifera L. sativa and sylvestris subspecies Ekhvaia et al., 2014;Salayeva et al., 2010), high levels of genetic diversity and heterozygosity (Arroyo-García et al., 2006;Imazio et al., 2013), and some particular alleles, scarcely represented in other genetic pools (Riaz et al., 2018). All these features are commonly found in crop domestication centers. Consistent with this evidence, archaeological and archeobotanical findings indicate that viticultural activities first began in this region during the early Neolithic Period , the earliest evidence of winemaking activities in Armenia being traced back to around 4,000 BCE (Barnard et al., 2011).
Nowadays, viticulture is one of the leading branches of the Armenian agricultural sector, accounting for more than 16,500 hectares of vineyards, which mainly produce grapes for wine production. According to the Vine and Wine Foundation of Armenia (https://vwfa.am/), the Armenian grape and wine sector produced 264,000 tonnes of table and wine grapes in 2020, and more than 11 M litres of wine. More than 80 % of grape production took place in the Armavir, Ararat and Aragatsotn provinces. The presence of abundant wild grapevine populations and the Armenian edafoclimatic conditions, altitudinal variation, isolated valleys and soil types likely prompted the generation of highly diverse grape varieties that are adapted to the local conditions (Dallakyan et al., 2020), which have been exploited for grape cultivation and winemaking in Armenia until today (Barnard et al., 2011). Nowadays, some of these varieties are exclusively found in private vineyards or gardens, where owners preserve old vines for their own consumption. Other cultivars can still be found in old viticultural areas and are at risk of extinction (Dallakyan et al., 2020;Margaryan et al., 2019;Nebish et al., 2017). Attempts have been made to minimise the loss of these local genetic resources, including the establishment of a series of grapevine collections. The first Armenian grapevine collection was established in 1950 by the Armenian Scientific Research Institute of Viticulture, Winemaking and Fruit growing in Yerevan (FAO Institute Code: ARM 02); it held 850 accessions of local cultivars, interspecific bred hybrids, international cultivars and wild forms. After its closure in the early 1990s, many of these cultivars were irreversibly lost, and only around 140 varieties were stored in the three minor grapevine collections created at (i) the Institute of Botany of the National Academy of Sciences of Armenia in Yerevan (ARM005), (ii) the Scientific Center of Viticulture, Winemaking and Fruit growing in Yerevan (ARM 06), and (iii) the Armenian Academy of Viticulture, Wine-Making and Fruit-Growing in Nalbandyan (ARM011). In 2016, most of the genetic resources stored in the Nalbandyan grapevine collection were transferred to Echmiadzin to be part of the new Armenian Grape Collection of the Scientific Center of Agriculture (Margaryan et al., 2019).
During this process, grape varieties were named under different appellations, and while new names were established in some areas, the old ones remained in others (Dallakyan et al., 2020). Molecular genetic studies are useful for identifying possible synonyms (accessions with identical genetic profiles, but with different names), homonyms (accessions with identical names, but with different genetic profiles) and misnames (accessions of one cultivar registered under the name of another cultivar) in grapevine collections. Such information contributes to the true-to-type identification of local varieties, making it easier to register them in national databases and to study their genetic relationships with local and international cultivars, as has been proved in numerous studies Maul et al., 2015;Popescu et al., 2017). Nowadays, the combined use of nuclear Simple Sequence Repeat (SSR) and Single Nucleotide Polymorphism (SNP) markers has proved efficient in the genetic identification of grapevine varieties, as well as in the description of first degree genetic relationships between local and international cultivars (Cunha et al., 2020;Maraš et al., 2020;Zinelabidine et al., 2012). When this information is combined with chloroplast genotype profiling, the maternal lineage of any cultivar can be determined, thus enabling its historical origin to be tracked down (Arroyo-García et al., 2006). Despite recent efforts to characterise the genetic diversity of Armenian cultivars (Dallakyan et al., 2020;Margaryan et al., 2019;Nebish et al., 2017), local grapevine collections still hold a high number of synonyms, homonyms and misnames (Dallakyan et al., 2020), which hinders their efficient conservation, characterisation, evaluation and eventual utilisation. In addition, the genetic relationships between the Armenian cultivars have been little explored, and information regarding their pedigree and likely origin is scarce. Therefore, the aim of this study was to use SSR and SNP markers in order to carry out a genetic characterisation of a series of traditional cultivars, most of them stored in the International Academy of Viticulture and Winemaking in Nalbandyan and presumed to be autochthonous Armenian varieties. From the results, it was possible to identify almost all the analysed accessions, as well as provide pedigree information about some of them, thereby improving knowledge of the Armenian grapevine genetic pool.

Plant material
Thirty-seven grapevine accessions from the Grape Collection of the Armenian Academy of Viticulture, Wine-Making and Fruit-Growing in Nalbandyan, Armavir (FAO Institute Code: ARM011) and from a local private orchard (two samples: AM16 and AM17) were analysed (Table 1). Accessions in the Nalbandyan grape collection correspond to local cultivars traditionally grown in Armenia or bred cultivars from recent breeding programmes, except for one sample (AM37) which was originally collected as a feral vine. Young leaves from all accessions were collected in situ, dried in silica gel and stored at room temperature until DNA extraction.

DNA extraction and genotyping
Whole genomic DNA was extracted from approximately 100 mg of ground leaf tissue according to the methods described by Zinelabidine et al. (2010). DNA quality and quantity was evaluated using a NanoDrop spectrophotometer (Thermo Scientific, Wilmington, USA) and by visual inspection with lambda DNA on an ethidium bromide-stained agarose gel (0.8 %). All samples were then genotyped at seven SSR markers in a single multiplex polymerase chain reaction (PCR): VVMD5, VVMD7, VVMD27, VVMD32, VVS2, VrZAG62, and VrZAG79, including the reference set of six SSR markers recommended for grapevine identification (Maul et al., 2012;This et al., 2004). Multiplex PCR reaction was carried out with 5 ng of DNA, 0.07 μM of VVMD32, 0.10 μM of VrZAG62, 0.12 μM of VVS2, 0.15 μM of VVMD7 and VrZAG79, 0.35 μM of VVMD27 and 0.50 μM of VVMD5 primers using QIAGEN multiplex PCR kit (Qiagen, Hilden, Germany). The forward primer of each pair was fluorescently labelled with different dyes: 6-FAM for VVMD27 and VVMD5, VIC for VVMD32 and VrZAG62, NED for VVS2 and VVMD7, and PET for VrZAG79. Amplification reactions were performed in a Thermal Cycler T100 (Bio-Rad, Hercules, USA) using the following PCR cycle: initial denaturation at 95 °C for 15 min, followed by 30 cycles of 95 °C for 45 s, 55 °C for 60 s, and 72 °C for 30 s, and a final extension step at 72 °C for 1 h. Sample AM23 was genotyped for one additional SSR marker (VVMD28), and samples AM08 and AM25 for four additional SSRs (VrZAG112, VrZAG29, VrZAG67 and VrZAG83), following the methods detailed in Ibáñez et al. (2009). DNA amplification products were mixed with 20 μl of highly deionized (Hi-Di) formamide, and 0.2 μl of GeneScan-500 LIZ size standards (both from Applied Biosystems, Foster City, CA, USA) and denaturalised at 95 °C for 5 min prior to capillary electrophoresis, which was performed in the genotyping platform of the Centro de Investigación Biomédica de La Rioja (CIBIR). Fragment sizes were determined by means of GeneMapper v.4.1 (Applied Biosystems, Darmstadt, Germany). In each analysis, one Tempranillo Tinto DNA was included as a positive control.
In parallel, DNAs were profiled for a set of 240 nuclear SNP markers applying Fluidigm technology, through the genotyping services provided by the Sequencing and Genotyping Unit of the Universidad del País Vasco (UPV/EHU). This set includes a core set of 48 SNPs used for cultivar identification and an extended set of 192 SNPs for parentage and diversity analyses (Cabezas et al., 2011;Cunha et al., 2020;Lijavetzky et al., 2007). The core set of 48 SNPs contained three chloroplast SNPs for the identification of the four most common grapevine chloroplast haplotypes (A, B, C and D) (Arroyo-García et al., 2006).

Variety identification
Non-redundant 48-SNP genetic profiles were compared pair-wise with those stored in the ICVV-SNP database for cultivar identification.
To date, this database contains more than 3,000 unique 48-SNP genotypes. In parallel, non-redundant SSR genetic profiles were compared with those stored in the VIVC (Maul, 2020), the European Vitis Database (Maul et al., 2012) and in the Armenian Vitis Database (Margaryan et al., 2019) for cultivar identification within a broader context.

Phylogenetic and parentage analyses
An Unweighted Neighbor Joining (UwNJ) distance tree was calculated to explore the relationship between the Armenian cultivars considered in this work and a group of 27 traditional grapevine cultivars from ten European countries: Austria (2 cultivars), France (5), Germany (2), Italy (5), Moldova (1), Montenegro (2), Portugal (2), Serbia (1), Slovenia (2) and Spain (5) (Supplementary Table 1). The 240-SNPs genetic profiles of these 27 cultivars were retrieved from the ICVV-SNP database. For this analysis, a dissimilarity matrix with 10,000 bootstrap steps was calculated using the DARwin software package v. 6.0.21 (Perrier and Jacquemond-Collet, 2006), discarding those SNPs with any missing data (90 SNPs were removed). This matrix was used to construct an UwNJ distance tree based on 1,000 bootstrap replicates. Similarly, another UwNJ distance tree was constructed for the Armenian cultivars analysed in this study following the same procedure (81 SNPs were removed), using VIVC information of cultivar main use (wine/table/ multi-purpose) for graphical representation.
For parentage analyses, the non-redundant 240-SNPs genotypes obtained in this study were merged with those stored in the ICVV-SNP database to detect possible first-order kinship relationships (trios and duos or parent-offspring pairs), using the likelihood-based method implemented in Cervus v.3.0 (Kalinowski et al., 2007) as previously described (Cunha et al., 2020). The likelihood of each detected trio and duo was evaluated taking into account the natural logarithm of the overall likelihood ratio (LOD) score. The maximum number of mismatching loci for trios and duos was set to 2 and 1 respectively, and only duos with LOD > 25 were considered. For each trio, chloroplast haplotype information was used to determine which of the putative parents acted as mother, according to the maternal transmission of chloroplasts in grapevine (Arroyo-García et al., 2006).

RESULTS
The combination of SNP and SSR genotyping of the 37 accessions explored in this work revealed 27 different genetic profiles (Table 2 and  Supplementary Table 2). Twenty-four grapevine cultivars were identified via the parallel comparison of the 48-SNP genetic profiles with those stored in the ICVV-SNP database, and the 7-SSR genetic profiles with international databases; the genetic profile of three accessions (AM16, AM17 and AM37), however, did not match the genetic profile of any previously registered grapevine varieties (  Table 2). For these 14 SNPs, AM25 was found to be homozygous for one of the two different alleles that were present in a heterozygous manner in AM08. Given their different SNP genetic profiles, these two accessions can be considered as two different cultivars.
As can be seen in  Belyi', and 'Tozot', all of them found twice. Regarding their main use, eight of the identified genotypes are registered in the VIVC database as table grape cultivars, nine as wine grape cultivars and six as cultivars with multiple uses. Chlorotype C was found to be the most abundant within the different genotypes analysed (18 times, 66.6 %), followed by chlorotype D (8 times, 29.6 %) and chlorotype B (once, 3.7 %) ( Table 2).
The UwNJ distance tree constructed from the 27 unique genetic profiles identified in this work and 27 profiles from autochthonous grapevine cultivars from diverse European countries grouped the genotypes into two clusters that clearly reflected their origin ( Figure 1A). The UwNJ tree constructed exclusively with the 27 unique genetic profiles identified in this study generated two main clusters ( Figure 1B), which mostly reflected the main use of the cultivar, as recorded in the VIVC database. Thus, one cluster grouped most of the varieties with a clear aptitude for winemaking (such as the white-berried cultivar 'Areni Spitak', or the black-berried cultivars 'Khardji Sev' and 'Tozot'), whilst the other grouped most of those suitable for table grape production (such as the cultivars 'Hayastan', 'Hayreniq' and 'Masis').
Parentage analysis based on SNP data was useful for revealing the full pedigree (trio) of eight of the genotypes identified in this work (Table 3), as well as eight first-order genetic relationships ( Table 4). The full pedigrees identified for 'Kishmish Khishrau', and 'Muscat Ottonel' had been already reported and confirmed in the literature (Lacombe et al., 2013), but it is the first time that the remaining six pedigrees (either previously reported in the literature or not) are If available, the VIVC variety number is given between brackets. If not, the SNP-ICVV genotype code is provided.
*For Khusaine Belyi (VIVC: 6203), both parents could have acted as parent 1 (mother) or parent 2 (father  TABLE 4. Summary of the first-order genetic relationships (duos) detected by SNP analyses involving (at least) one of the genotypes of the Armenian samples studied in this work. If available, the VIVC variety number is given between brackets. If not, the SNP-ICVV genetic code is provided.
supported by molecular analyses. All pedigrees were fully confirmed by means of the SSR data available, except the one proposed for 'Khusaine Belyi', which has one allele for VVS2 (147) absent in its suggested two genitors ('Khusaine Belyi' is 147:155, whereas 'Dzhandzhal Kara' and 'Unknown' (AM37) are 125:155 and 143:155 respectively) (Supplementary Table  2) (Maul, 2020). Moreover, our results indicate a relevant role for the female cultivar 'Angur Kalan' in the genotypes identified here, being a genitor in four of the eight full trios identified in this study. In addition, the genotype obtained for 'Nerkeni' was found to be compatible with a self-cross of the genotype obtained for 'Charentsi'. Lastly, six new duos were identified within the genotypes explored in this study; as shown in Table 4, in many of them the cultivars involved were Armenian, indicating the close pedigree relationships that exist among local grapevines.

DISCUSSION
Ancient and almost neglected cultivars in the South Caucasus have been found to have unique traits for facing current viticulture challenges (Bitsadze et al., 2015;Sargolzaei et al., 2021); this highlights the relevance of exploring local grapevine germplasm as alternative genetic sources of variability for future breeding programmes.
The identification of local Armenian grapevines is the first step in their study and eventual use; however, recent reports have shown that many have not been properly identified (Dallakyan et al., 2020;Margaryan et al., 2019;Nebish et al., 2017). In the present study, the analysis of 37 grapevine local accessions by means of SNP and SSR markers resulted in the identification of 27 unique genotypes, of which 24 were already recorded in the VIVC database, most of them as Armenian cultivars (Table 1). Therefore, three new genetic profiles, which were not registered in the three international databases used in the study, were identified: 'Karmir Khach', 'Karmir Muscat' and 'Unknown'. These three samples were obtained from a private orchard ('Karmir Khach' and 'Karmir Muscat', both locally appreciated for fresh fruit production), and from the wild ('Unknown'). Although the latter sample was collected in the wild, it is unlikely that it is a sylvestris plant, since in a preliminary genetic structure analysis between European sativa and sylvestris genotypes it clustered with the sativa individuals (data not shown). In addition, it was identified as one of the potential genitors of the cultivar 'Khusaine Belyi' (Table 3), so it is more likely that this 'Unknown' genotype corresponds to an ancient cultivar in traditional Armenian viticulture. The parent-offspring relationship observed between 'Khusaine Belyi' and its other potential genitor ('Dzhandzhal Kara') has already been suggested by Lacombe et al. (2013) and Laucou et al. (2018). Because the three involved genotypes carried the chlorotype C (Table 3), it was not possible to determine which one acted as female genitor in this cross.
Parentage analyses confirmed the pedigrees previously proposed for the cultivars 'Kishmish Kishrau' ('Angur Kalan' × 'Kishmish Chernyi'), and 'Muscat Ottonel' ('Ingram's Muscat' × 'Chasselas Blanc') by means of SSR profiling (Lacombe et al., 2013), and they supported the pedigree indicated for the bred cultivar 'Hayastan' ('Angur Kalan' × 'Italia'), for which no previous molecular data was available . In addition, our results were useful for clarifying the pedigree of the Armenian cultivar 'Hayreniq'. As declared by the breeders, 'Hayreniq' is an offspring of a cross between 'Italia' and 'Muscat Hamburg' cultivars (Golodriga et al., 1984), but this pedigree is not compatible with the SSR data stored in the VIVC database (Maul, 2020). Our SSR and SNP results strongly indicate that 'Hayreniq' resulted from a cross between the cultivars 'Angur Kalan' and 'Alphonse Lavallee'. The female cultivar 'Angur Kalan' (syn. 'Nimrang') was found to be a major founder of local Armenian genetic resources (Table 3), which agrees with its common use as a female genitor in breeding programmes aimed at enhancing the productivity of vineyards during the development of viticulture in the South Caucasus (Maghradze et al., 2020). For practical reasons, renowned breeders like Bruno Bruni, Giovanni Dalmasso, Pierre Landot or Alberto Pirovano also recurrently used female cultivars as genitors in their breeding programmes. As a result, female-flowered cultivars like 'Bicane' or 'Chaouch Blanc' figure as progenitors of numerous cultivars bred during the nineteenth and the beginning of the twentieth centuries, especially of table grapes (Lacombe et al., 2013;Vargas et al., 2009). Furthermore, the results of the present study, indicate that 'Nerkeni' (AM25) may be the result of the self-pollination of 'Charentsi' (AM08). Although hermaphrodite V. vinifera genotypes can theoretically be selfed, this is not a common mechanism for creating new cultivars in V. vinifera, and only few examples of cultivated varieties derived from self-pollination events have been molecularly confirmed (such as 'Covè' or 'Čubrica-2', selfings of 'Harslevelu' and 'Čubrica' respectively (Cipriani et al., 2010;Maraš et al., 2020)). In fact, selfed offsprings are at a disadvantage, because inbreeding exposes the harmful nature of deleterious mutations hidden in heterozygous state (Zhou et al., 2019). Interestingly, the VIVC database includes a variety named 'Nerkeni' (VIVC variety code: 8489), but unfortunately SSR information is lacking for a direct comparison to be made with the genetic profile obtained in this study. Nevertheless, the pedigree indicated by the breeders for this variety ('Saperavi' × Mixture of pollen Areni (Malan) + Kachet + Hybrid Nº24 (= Garandmak x Richter 31)) (Katarjan, 1962) is not compatible with the genetic profile obtained for 'Nerkeni'; for example, the proposed female genitor of 'Nerkeni' ('Saperavi') is 239:239 and 188:200 for VVMD7 and VrZAG62 respectively (Maul, 2020), whilst the profile obtained for 'Nerkeni' (AM25) was 247:249 and 194:196 for the same two loci (Supplementary Table 2). This suggests that (a) the pedigree data is wrong (as also observed for 'Hayreniq'), or (b) it is one case of homonymy, and the 'Nerkeni' studied here corresponds to a different cultivar with the same name.
In the present study, multiple cases of synonyms, homonyms and misnames were revealed. As an example of synonymy, up to six different accessions registered under diverse local names were genetically identified as 'Areni Sev' (Table 2). In some cases, these synonyms indicate some particular phenotypic features of the accession, like the one observed for the 'Areni Sev' accession registered as 'Lyustra-2'. Lyustra means 'chandelier' in Russian, which reflects the singular loose cluster architecture of this 'Areni Sev' accession. The high prevalence of 'Areni Sev' within the set of accessions analysed in this work agrees with the high relevance of this cultivar for Armenian viticulture, which has been widely used for red wine elaboration in different Armenian regions for centuries (Dallakyan et al., 2020). Other cases of synonymy were detected for the cultivars 'Arakseni Chernyi' (AM29), 'Kishmish Khishrau' (AM15), 'Tozot' (AM32), and 'Voskeat' (AM18), which were registered under the local names 'Sev Araqseni', 'Kaqavik', 'Sali', and 'Kharch Kharji' respectively. Interestingly, the cultivar 'Muscat Ottonel' was found in the grape collection under the local name 'Muscat Spitak' (AM23), despite this term being registered as a synonym for the cultivar 'Muscat a Petits Grains Blancs' (Maul, 2020). Regarding the detected homonyms, accessions registered as 'Avagi' were identified as 'Areni Sev', 'Eghegnadzori sev' and 'Hakobi Vordi', and accessions named 'Lyustra' were identified as 'Areni Sev' and 'Khardji sev'.
On the other hand, two accessions (AM26 and AM30, 'Parvana' and 'Shahumyani' respectively) were identified as 'Khusaine Belyi', a white-berried table cultivar existing in numerous grapevine collections under many synonyms (Maul, 2020) and progenitor of other table grape cultivars (Lacombe et al., 2013). Accession AM30 is seeded (Table 1), which agrees with the seed phenotype recorded for 'Khusaine Belyi' in the VIVC database (Maul, 2020). Interestingly, accession AM26 has been previously described as stenospermocarpic , suggesting that it could be a 'Khusaine Belyi' seedless somatic variant. If confirmed, the analysis of these two accessions could be of high interest for determining the genetic and molecular causes of this seedless phenotype, and thus providing new insights into this relevant trait for table grape breeding, as recently done with other variant pairs (Costantini et al., 2021;Royo et al., 2016;Royo et al., 2018). The seedless accession AM26 was found to be wrongly named 'Parvana', which is another Armenian whiteberried seedless cultivar with a similar cluster phenotype to that of AM26, what could be the origin of the misnaming. Unfortunately, this was not the only misnamed accession, as the accessions wrongly named under the cultivar names 'Ararati' (AM01), 'Kakhet' (AM14), 'Mskhali' (AM22), 'Qrdi Khaghogh' (AM27), and 'Vani' (AM34), were found to correspond to the cultivars 'Karmir Kakhani', 'Areni Sev', 'Masis', 'Areni Spitak' and 'Hayastan' respectively (Table 2).
Lastly, the studied genotypes showed a clear genetic differentiation from the set of Western and Central European varieties used as a reference ( Figure 1A), and they had some microsatellite alleles poorly represented in other genetic pools (like the 292 allele detected in VVMD32) -which are two of the expected features of a putative domestication centre (Sargolzaei et al., 2021). These results are in line with the generally accepted assumption of the South Caucasus as the cradle of grapevine domestication, which has been widely supported by genetic, archaeological and cultural evidence (Arroyo-García et al., 2006;Bouby et al., 2021;McGovern et al., 1997;Riaz et al., 2018). In addition, we found a predominance of chlorotypes C and D in the identified genotypes, which is commonly observed among cultivars from Near East and Middle East regions (Arroyo-García et al., 2006).

CONCLUSIONS
Our results help better understand the diversity of the Armenian grapevine germplasm, providing useful information for its correct identification.
Despite the reduced number of accessions analysed, several cases of synonyms, homonyms and misnames were revealed. The pedigrees of several Armenian varieties have been confirmed and some others proposed for the first time.
Other additional compatible parent-offspring relationships have been found, which need to be confirmed in future analyses. Furthermore, the potential finding of a novel seedless somatic variant highlights the importance of exploring the Armenian genetic pool, in order to study and exploit novel sources of genetic diversity for breeding novel cultivars which meet consumer expectations and current viticulture challenges. Unfortunately, the Nalbandyan Grape Collection of the Armenian Academy of Viticulture, Wine-Making and Fruit-Growing is no longer operational, but the majority of the accessions analysed in this study have been transferred to the Armenian National Grape Collection of the Scientific Center of Agriculture in Echmiadzin (Armavir, Armenia). Thus, the information provided in this study will be useful for curating and updating information about the grapevine germplasm preserved in this new collection.