Radish (Raphanus sativus L., 2n = 2x = 18) is an important root vegetable belonging to the Brassicaceae family. It is widely cultivated across the globe as a biennial or annual horticultural crop. Radish is primarily valued for its edible roots, leaves, and shoots, which are rich in nutrients (Banihani, 2017; Xie et al., 2018). The fleshy taproot contains various nutrients, including carbohydrates, minerals, phenolic compounds, dietary fiber, and glucosinolates (Manivannan et al., 2019; Gamba et al., 2021). Premature bolting triggered by low temperatures and intense light can adversely affect the yield and quality of the radish (Jung et al., 2016). From a breeding perspective, early bolting and flowering in radish can shorten the breeding cycle, thereby accelerating the development of superior varieties. Thus, precise regulation of bolting and flowering time in radish is crucial for increasing economic returns and breeding efficiency. However, because bolting and flowering are quantitative traits influenced by various environmental factors, QTL mapping is essential for identifying the key genes involved. Moreover, a complete genome sequence is required to establish an accurate genomic reference framework for effective QTL mapping.

The first radish genome was sequenced in 2014, and over the subsequent 6–8 years, several radish varieties have been sequenced (Jeong et al., 2014; Kitashiba et al., 2014; Mitsui et al., 2015; Zhang et al., 2015). With the advent of third-generation sequencing technologies, Shirasawa et al. (2020) used single-molecule real-time (SMRT) sequencing to generate a de novo assembly of the Japanese cultivar “Sakurajima Daikon”, resulting in a 504.5 Mb genome sequence. The contig N50 reached 1.2 Mb, and 69.3% of the assembly was mapped to the chromosomal level. Zhang et al. (2021) sequenced the genomes of 11 representative species, subspecies, and varieties within the Raphanus genus, constructing a multi-tiered, genus-level pan-genome. This work provided insights into whole-genome variation, gene flow, and evolutionary mechanisms among cultivated, wild, and weedy radishes. Xu et al. (2023) used a combination of Illumina, PacBio, and BioNano mapping technologies on the NAU-LB radish inbred line, assembling a reference genome of 476.32 Mb with a scaffold N50 of 56.88 Mb. Using Hi-C data, they anchored 448.12 Mb (94.08%) of the sequences to nine radish chromosomes and annotated 40,306 protein-coding genes.

However, previous assemblies of radish genomes contained multiple gaps, which impede downstream analyses such as gene function studies, variation detection, and other genomic investigations. Thus, a gap-free genome provides a more comprehensive and accurate genomic resource. The advent of third-generation sequencing technologies has enabled the generation of T2T chromosome sequences, facilitating the precise reconstruction of repetitive regions and revealing the centromere and telomere structures. This advancement provides more accurate genomic information to support genetic domestication, molecular breeding, and crop improvement. Numerous plant genomes, including those of Arabidopsis thaliana (Naish et al., 2021; Hou et al., 2022), Oryza sativa (Li et al., 2021), Zea mays (Chen et al., 2023), Brassica rapa (Zhang et al., 2023), Momordica charantia (Fu et al., 2023), Solanum lycopersicum (Wang et al., 2025), Phaseolus vulgaris L. (Zhao et al., 2025), and Firmiana hainanensis (Dong et al., 2025), have been assembled to near-complete or T2T levels using third-generation sequencing platforms. The development of high-throughput molecular markers that support genetic analysis, gene discovery, and breeding programs is one of the key outcomes of crop genome sequencing. Notably, a major advantage of genome sequencing-based marker development is the increased availability of QTL data, which is essential for understanding key agronomic traits in crops (Khan et al., 2021; Tang et al., 2021). For example, Wei et al. (2023) compiled 1294 QTLs from 67 studies and mapped them onto a T2T genome.

Bolting and flowering are critical transitional stages in plant growth and development, marking the transition from the vegetative to the reproductive phase. Studies have shown that plants regulate these processes by sensing and integrating external environmental cues and endogenous signals. These signals operate through six major genetic pathways: vernalization, photoperiod, gibberellin, temperature, age, and autonomous pathways. Together, these pathways regulate core flowering genes such as FLOWERING LOCUS T (FT), SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1), FLOWERING LOCUS C (FLC), and LEAFY (LFY), thereby promoting or suppressing bolting and flowering phenotypes (EmamiKempken, 2019; He et al., 2019). However, the molecular mechanisms underlying bolting and flowering in radish remain largely unclear. Transcriptomic data are frequently used to identify key genes for molecular breeding, for example, in the discovery of cuticle biosynthesis genes in Hylocereus undatus (García-Coronado et al., 2024), transcription factors and genes regulating Quercus acutissima Carruth development (Byeon et al., 2024), and candidate genes associated with the flowering time of Brassica napus (Song et al., 2021). Therefore, integrating high-throughput sequencing technologies with phenotypic data under diverse environmental conditions is essential for effective QTL mapping and transcriptome analysis to elucidate the gene networks regulating bolting and flowering in radish.

In this study, two high-generation autogamous radish lines, differing in bolting time by approximately 120 days, were selected as parental materials to develop a near-isogenic line (NIL), an NIL-F2 mapping population, and an NIL-F2:3 family population for phenotypic evaluation. High-quality T2T genomes of the late-bolting parent line C60213 were assembled using PacBio and Oxford nanopore (ONT) sequencing, successfully closing all gaps present in existing reference genomes. Using the genome assembly and annotation of C60213, along with a genetic map and phenotypic data from the NIL-F2:3 family and its parents under diverse environmental conditions, we performed QTL mapping for bolting and flowering traits in radish. A total of eight QTL intervals associated with these traits were identified. By integrating transcriptomic data across different developmental stages, we performed candidate gene screening and identified MYO-INOSITOL-1-PHOSPHATE SYNTHASE 3 (RsMIPS3) as a key regulator of bolting and flowering in radish. The function of RsMIPS3 was validated by its overexpression in Arabidopsis. These findings provide valuable genomic resources for radish genome research and offer a reference framework for exploring bolting and flowering traits in radish and other Brassicaceae species, as well as for developing related molecular markers.