*2.2. Identification of SSRs*

The MISA (Microsatellite) tool was used to identify SSR loci in the transcriptome data, and a total of 15,384 SSRs were identified from 98,389 unigenes; the SSR frequency in the transcriptome was 15.63%, and the mean distance of the SSRs in the unigenes was 6.39 kb. It is well known that the distribution frequency of SSR loci varies greatly between different species, which may be related to the genome size of species, the organ and period of the sampled plant used for transcriptome sequencing, the SSR development methods, and the screening criteria [30]. Species with higher frequencies include, for instance, hybrid *Cymbidium* (58.64%) [31], *Morus alba* L. (45.56%) [32], and *Bougainvillea glabra* (44.91%) [33]. Species with lower frequencies include *Paeonia suffruticosa* (6.4%) [34], *Pinus elliottii* (4.80%) [35], and *Chrysanthemum morifolium* (2.84%) [36]. Kumpatal [37] found that the distribution frequency of SSR loci in dicotyledons ranged from 2.65% to 16.82%, with an average of 9.73%. The distribution frequency of SSR loci in the flower transcriptome of *L. yunnanensis* is consistent with the distribution frequency range of SSR loci in dicotyledons, and higher than the average. Therefore, the flower transcriptome data of *L. yunnanensis* had abundant SSR loci, and could be used for the development of SSR markers.

There were abundant SSR types in the flower transcriptome of *L. yunnanensis*, with the distribution of mononucleotide to hexanucleotide repeat types, but the proportion of different repeat types was significantly different. The number of mononucleotide repeat motifs found to be the most frequent was 8889, accounting for 60.29% of the total repeat motifs. Second, dinucleotide and trinucleotide repeat motifs were 3340 and 2309, accounting for 22.65% and 15.66%, respectively, of the total repeat motifs. The tetranucleotide, pentanucleotide, and hexanucleotide repeat motifs accounted for only 158, 23, and 25, respectively. In this study, the number of short-repeat motifs (mononucleotide, dinucleotide, and trinucleotide) was more than that of long-repeat motifs, which is consistent with the distribution in most species, and supports the view that long-repeat motifs have high variability [38]. As shown in Table 3, the mono-, di-, tri-, tetra-, penta-, and hexanucleotide motifs with the highest frequency were A/T, AG/CT, AAG/CTT, AAAG/CTTT, CCTTC/GAAGG, and GAGAC/GTCTC, respectively. The frequency of mononucleotide repeats was the highest. This result was similar to those for *Ricinus communis* [39], *Glycyrrhiza uralensis* Fisch [40], *Arachis hypogaea* L. [41], and *Elaeagnus mollis* Diels [24], which showed that the A/T repeat motif accounted for the largest proportion of SSRs. The proportion of 12–21 bp repeats in SSRs under different repeat types and repeat times was more than 95%, among which the proportion of 5 times and trinucleotide repeat motifs was the largest (22.25%), followed by that of 6 times and dinucleotide repeat motifs (16.48%) (Table 4).


**Table 3.** Frequency of different repeat motifs in the SSRs of the *L. yunnanensis* flower transcriptome.

**Table 4.** Number of SSRs of different repeat motifs on different repeat times in the flower transcriptome of *L. yunnanensis*.


The polymorphism level of SSR markers is an important basis for evaluating their availability, and the length of SSRs is an important factor affecting their polymorphism. The results showed that the polymorphism was high when the length of SSRs was more than or equal to 20 bp, medium when the length was 12–19 bp, and extremely low when the length was less than 12 bp [42]. Therefore, more than 95% SSRs were had a moderate-to-high level of polymorphism in this study. In addition, lower-grade SSR motifs (i.e., mononucleotide, dinucleotide, and trinucleotide) tend to produce polymorphism more easily than higherorder SSR motifs [43,44]. Due to the large number of mononucleotides, dinucleotides, and trinucleotides in this study, it can be predicted that SSRs obtained from the *L. yunnanensis* flower transcriptome have a high potential for polymorphism, and this leads to a high application value in the study of its molecular markers.
