Vigna unguiculata (L.) Walp. is a dryland legume crop, providing essential food and nutritional security for millions of people across the semi-arid tropics, in Africa, Asia and Latin America. However, as a typical 'orphan crop', cowpea has long remained underrepresented in global genomic research to support crop improvement. Here, we conducted the largest genetic diversity analysis of cowpea to date, comprising 10 617 accessions sourced from seven international collections. Using genotyping-by-sequencing, we characterised the global patterns of genetic diversity, assessed redundancy within and across collections, and examined the geographic structure of the cowpea global allele pool. Our results revealed nine distinct genetic groups with clear geographic associations and fine-scale population differentiation, reflecting dispersal history, regional adaptation and the influence of modern breeding. Duplication across collections was detected, highlighting the need for improved curation and integration of germplasm resources. Landraces from sub-Saharan Africa do not fully capture the genetic diversity present in several other geographic regions, indicating the existence of abundant and untapped genetic resources worldwide. These findings not only provide insights into the genetic structure and evolutionary history of cowpea but also offer a valuable foundation for harnessing global germplasm diversity to enhance breeding potential and accelerate crop improvement.

Cowpea (Vigna unguiculata L. Walp.) is an Africa-originated legume providing a vital source of protein for millions across resource constrained areas of Africa, Asia and Latin America. Understanding genetic diversity in cowpea is pivotal to its genetic improvement. Here, we assemble 20 high-quality genomes representing cowpea global diversity and construct a graph-based pan-genome to characterize genetic variation. The pan-genome comprises 29,557 orthogroups, with approximately 73% conserved across all accessions. Extensive structural variations (SVs) are uncovered, affecting thousands of coding sequences and gene expression. Thousands of SVs are found under selection between two cowpea subspecies, highlighting their roles in subspecies divergence. We pinpoint key SVs underlying variation in pod length and seed number per pod. Two SV clusters appear to modulate pod length through altering expression and function of VuWAK and VuGA2ox2 genes. These findings provide insights into cowpea genetic diversity and establish a fundamental resource for cowpea genetic improvement.

Background Human interaction with plants reshapes genetic diversity across species and lineages, yet this process remains poorly understood in perennial woody plants cultivated for non-fruit products. Cinnamon, primarily derived from Cinnamomum cassia , represents a compelling subject to investigate how ecological adaptation and long-term cultivation have shaped its genomic diversity and driven variation in its key bioactive compounds, namely cinnamaldehyde and its derivatives. Results To investigate the divergence of Chinese cinnamon ( C. cassia var. cassia ) and Vietnamese cinnamon ( C. cassia var. macrophyllum ), we integrated a de novo genome assembly, whole-genome resequencing of 98 accessions, and metabolomic and transcriptomic profiling of multiple tissues of both varieties. Phylogenomic and demographic analyses revealed that the two varieties diverged prior to documented human utilization. Vietnamese cinnamon formed a basal lineage with high genetic diversity, whereas Chinese cinnamon exhibited reduced genetic diversity and weak population structure, likely resulting from prolonged intensive cultivation. Genome-wide selection scans indicated genes involved in phenylpropanoid biosynthesis were under divergent selection. Metabolomic and transcriptomic profiling uncovered differentiation in cinnamaldehyde derivatives across tissues and varieties and pinpointed candidate enzymes in their biosynthesis. Functional validation demonstrated that two O-methyltransferases, Ccas10G000326 and Ccas05G001459, effectively enhance the transformation of 2-hydroxycinnamaldehyde into 2-methoxycinnamaldehyde. Conclusions Our findings demonstrate that persistent cultivation of a limited gene pool diminishes genetic diversity in perennial trees, whereas lineages under lower utilization pressure preserve greater evolutionary potential. It highlights that conserving and integrating underutilized genetic resources is essential for the sustainable use and improvement of woody crops.

Nucleotide-binding leucine-rich repeat (NLR) genes constitute one of the largest families of plant immune receptors and are central to crop disease resistance. Rice (Oryza sativa L.), maize (Zea mays L.), and sorghum (Sorghum bicolor (L.) Moench) are major crops underpinning global food security, and enhancing their immunity is critical for yield stability. Here, we conducted a comprehensive analysis of NLR repertoires across 75 genomes from these three grass crops, identifying 24,944 genes that exhibit extensive intra- and interspecific variation. The crops share broadly comparable domain architectures of NLR genes, despite their differences in gene number and genomic clustering pattern. We detected 68 integrated domains, revealing high diversity and pronounced lineage specificity. The pan-NLR profiles of the three crops were generally similar, with strong lineage-restricted expansion of NLR gene families. Cross-species comparisons revealed distinct proportions of common gene and lineage-specific genes between the crops, suggesting divergent evolutionary trajectories. Widespread presence/absence variation of NLR genes was identified within each crop, with an enrichment in clustered loci. NLRs, especially clustered ones, were significantly enriched in disease-resistance quantitative trait locus (QTL) hotspots, supporting their critical role in crop disease resistance. Notably, conserved Rp1-D clusters co-localized with resistance QTLs in the three crops, underscoring functional conservation and the translational potential of cross-species NLR studies. Collectively, these findings provide new insights into the genetic diversity and evolutionary dynamics of NLRs in three major grass crops and provide a valuable genomic resource for enhancing crop disease resistance.

Abstract Callicarpa nudiflora is a medicinal plant rich in phenylethanoid glycosides (PhGs), particularly verbascoside, but the genetic basis of PhG biosynthesis and natural variation in its content remain largely unexplored. Here, we assembled a chromosome-level genome of C. nudiflora (511 Mb, contig N50 = 12.78 Mb) with 39 317 annotated protein-coding genes and 99.5% BUSCO completeness and quantified its bioactive compounds and gene expression across multiple tissues. Using the dataset, we elucidated the complete biosynthetic pathway of verbascoside and demonstrated multiple paralogous enzymes catalyzing one step, a functional redundancy that may contribute to the high verbascoside accumulation observed in this species. Population genomic analysis, together with quantification of verbascoside content of 95 samples grown in a common garden over three consecutive years, revealed that Southcentral Hainan (HN1) was identified as the genetic diversity center and exhibited the the highest content of verbascoside. Selective sweep analysis comparing high- and low-content individuals identified CnUGT79G7, a key glycosyltransferase in the verbascoside pathway, as being under positive selection. Two major haplotypes of CnUGT79G7 were identified: Hap1, a high-expression haplotype prevalent in the top-geoherb region (77% frequency), and Hap2, a low-expression haplotype fixed in other populations. Although the two haplotypes encode proteins with similar catalytic activity, Hap1 is associated with significantly higher CnUGT79G7 expression and elevated verbascoside content, suggesting that cis-regulatory variation underlies chemotypic differentiation. Collectively, this work identifies key molecular components for verbascoside biosynthesis, establishes a genomic and metabolic framework for understanding top-geoherbalism, and provides valuable resources for the conservation and molecular breeding of medicinal plants.

Cowpea (Vigna unguiculata) is a versatile legume crop providing a critical source of grain, vegetable and forage globally. Cultivated cowpea is classified into two main subspecies, subsp. sesquipedalis for fresh-pod vegetable and subsp. unguiculata for grain production. Here, we present two complete telomere-to-telomere (T2T) assemblies for the grain-type inbred lines HJD and vegetable-type FC6 through integrating PacBio HiFi reads, Oxford Nanopore ultralong reads, and Hi-C data. The T2T genomes demonstrated improved contiguity, completeness, and accuracy compared to existing genomes, revealing clear telomeric and centromeric features. Comparative analysis of the T2T genomes highlighted inversions underlying subspecies divergence in cowpea. Evolutionary analysis uncovered contraction of gene families related to symbiosis in HJD, consist with its reduced root nodules compared to FC6. Distribution and composition of tandem repeat arrays and transposable elements in centromeric regions were largely conserved in cowpea, but displayed pronounced variation among Phaseoleae. Furthermore, frequent shifts of centromeric locations coincided with inversions found in Phaseoleae. Overall, this study provides a set of fundamental resources for cowpea improvement and enhances our understanding of cowpea subspecies divergence and genome evolution in Phaseoleae.

Furocoumarins (FCs), important natural compounds with biodefense roles and pharmacological activities, are notably abundant in medicinal plant Angelica dahurica. However, its accumulation patterns over development stages in FC-enriched tissue, biosynthetic pathways, and regulatory mechanisms in A. dahurica remain elusive. Here, we quantified the concentration dynamics of 17 coumarins across six developmental stages of root and found a gradual decrease in FC concentration as the roots develop. Using a de-novo assembled chromosome-level genome for A. dahurica, we conducted integrative multi-omics analyses to screen out candidate genes to fill in the sole missing step in the biosynthesis of imperatorin and isoimperatorin. This revealed that CYP71AZ18 catalyzes hydroxylation at the C-5 position of psoralen to generate bergaptol, while CYP71AZ19 and CYP83F95 catalyze hydroxylation at the C-8 position to produce xanthotoxol, notably indicating that a single step is catalyzed by two genes from distinct CYP450 subfamilies in this species. CYP71AZ19 originated from a proximal duplication event of CYP71AZ18, specific to A. dahurica, and subsequently underwent neofunctionalization. Accessible chromatin regions (ACRs), especially proximal ACRs, correlated with high gene expression levels, and the three validated genes exhibited strong signals of ACRs, showing the importance of chromosomal accessibility in regulating metabolite biosynthesis.

Cowpea ( Vigna unguiculata L. Walp) is a dryland legume crop, providing essential food and nutritional security for millions of people across the semi-arid tropics, in Africa, Asia, and Latin America. However, as a typical “orphan crop”, cowpea has long remained underrepresented in global genomic research to support crop improvement. Here, we conducted the largest genetic diversity analysis of cowpea to date, comprising 10,617 accessions sourced from seven international collections. Using genotyping-by-sequencing, we characterized the global patterns of genetic diversity, assessed redundancy within and across collections, and examined the geographic structure of the cowpea global allele pool. Our results revealed nine distinct genetic groups with clear geographic associations and fine-scale population differentiation, reflecting dispersal history, regional adaptation and the influence of modern breeding. Duplication across collections was detected, highlighting the need for improved curation and integration of germplasm resources. Landraces from sub-Saharan Africa do not fully capture the genetic diversity present in several other geographic regions, indicating the existence of abundant and untapped genetic resources worldwide. These findings not only provide insights into the genetic structure and evolutionary history of cowpea but also offer a valuable foundation for harnessing global germplasm diversity to enhance breeding potential and accelerate crop improvement.

The evolutionary mechanisms underlying ecological divergence between closely related species remain a central question in biology. Scirpus mariqueter is a coastal halophyte thriving in the saline intertidal zone and exhibits marked adaptive differences compared to its freshwater relative Bolboschoenus planiculmis. However, the genomic and physiological bases of its salt tolerance remain poorly understood. We generated high-quality genome assemblies for both species and investigated the anatomical and physiological innovations underpinning S. mariqueter's adaptation to extreme environments. Morphological analyses revealed that S. mariqueter evolved specialized traits-including denser leaf palisade tissues, enhanced stem aerenchyma, and compact root cortices-synergistically limiting salt intrusion. Using chromosome-level genomes, we identified lineage-specific expansions in S. mariqueter of gene families critical for salinity tolerance, including those regulating carbohydrate metabolism, photosynthetic fidelity, and reactive oxygen species (ROS) detoxification. Strikingly, germin-like protein (GLP) and wound-induced protein (WIP) families contain tandem repeats mediating ROS scavenging and cell wall integrity, underwent adaptive expansion, paralleling anatomical innovations. Physiological profiling under salt stress confirmed S. mariqueter's unique capacity to maintain photosynthetic activity and carbohydrate production, directly linking genomic adaptations to functional resilience. This study reveals an adaptive strategy whereby structural modifications, diversification of stress-responsive gene families, and metabolic stability collectively enable S. mariqueter to thrive in saline ecosystems.

Flavonoids have been shown to play an essential role in plant growth and fertility. 4-Coumarate CoA ligase (4CL) is one of the indispensable enzymes involved in the biosynthesis of flavonoids. However, the role of 4CL and flavonoids in impact on cotton fertility is still unknown. In this study, on the basis of identification of an additional Gh4CL gene, Gh4CL20A, by using an updated G. hirsutum genome, we found that Gh4CL20A and its homologous Gh4CL20 were preferentially expressed in petals and stamens. The petals of the loss-of-function Gh4CL20/Gh4CL20A mutant generated by CRISPR/Cas9 gene editing remained white until wilting. Notably, the mutant showed indehiscent anthers, reduced number of pollen grains and pollen viability, leading to male sterility. Histological analysis revealed that abnormal degradation of anther tapetum at the tetrad stage and abnormal pollen grain development at the mature stage caused male sterility of the gene editing mutant. Analysis of the anther transcriptome identified a total of 10574 and 11962 genes up- and down-regulated in the mutant, respectively, compared to the wild-type. GO, KEGG, and WGCNA analyses linked the abnormality of the mutant anthers to the defective flavonoid biosynthetic pathway, leading to decreased activity of 4CL and chalcone isomerase (CHI) and reduced accumulation of flavonoids in the mutant. These results imply a role of Gh4CL20/Gh4CL20A in assuring proper development of cotton anthers by regulating flavonoid metabolism. This study elucidates a molecular mechanism underlying cotton anther development and provides candidate genes for creating cotton male sterile germplasm that has the potential to be used in production of hybrid seeds.

Homosporous lycophytes (Lycopodiaceae) are a deeply diverged lineage in the plant tree of life, having split from heterosporous lycophytes (Selaginella and Isoetes) ~400 Mya. Compared to the heterosporous lineage, Lycopodiaceae has markedly larger genome sizes and remains the last major plant clade for which no chromosome-level assembly has been available. Here, we present chromosomal genome assemblies for two homosporous lycophyte species, the allotetraploid Huperzia asiatica and the diploid Diphasiastrum complanatum. Remarkably, despite that the two species diverged ~350 Mya, around 30% of the genes are still in syntenic blocks. Furthermore, both genomes had undergone independent whole genome duplications, and the resulting intragenomic syntenies have likewise been preserved relatively well. Such slow genome evolution over deep time is in stark contrast to heterosporous lycophytes and is correlated with a decelerated rate of nucleotide substitution. Together, the genomes of H. asiatica and D. complanatum not only fill a crucial gap in the plant genomic landscape but also highlight a potentially meaningful genomic contrast between homosporous and heterosporous species.

Evolutionary dynamics of inversion and its impact on biochemical traits are a puzzling question. Here, we show abundance of inversions in three Curcuma species (turmeric, hidden ginger and Siam tulip). Genes within inversions display higher long terminal repeat content and lower expression level compared with genomic background, suggesting inversions in Curcuma experience relaxed genetic constraints. It is corroborated by depletion of selected SNPs and enrichment of deleterious mutations in inversions detected among 56 Siam tulip cultivars. Functional verification of tandem duplicated terpene synthase (TPS) genes reveals that genes within inversions become pseudogenes, while genes outside retain catalytic function. Our findings suggest that inversions act as a counteracting force against tandem duplication in balancing birth and death of TPS genes and modulating terpenoid contents in Curcuma. This study provides an empirical example that inversions are likely not adaptive but affect biochemical traits. Evolutionary dynamics of inversions and their impact on biochemical traits are unclear. Here, the authors report the genome assemblies of three Curcuma species and find that terpene synthase genes quite often become pseudogenes inside an inversion, indicating inversions act as a counteracting force against gene tandem duplication.

Furocoumarins (FCs) are crucial natural products playing a dual role as plant defense molecules and pharmacologically active substances. Angelica dahurica is a renowned herb with diverse and abundant FCs. However, the accumulation pattern over developmental stages, biosynthesis pathway and regulatory mechanisms of FCs in A. dahurica remain elusive, hindering the production of FCs via synthetic biology approaches. Here, we constructed a chromosome-level reference genome for A. dahurica and quantified the content dynamics of 17 coumarins across six developmental stages of its medicinal organ, root. It showed a gradual decrease in FC concentration with root enlargement. The combined analyses of transcriptomic and metabolomic data, together with in vivo enzymatic assay, confirmed that CYP71AZ18 was involved in the biosynthesis of bergaptol, whereas CYP71AZ19 and CYP83F95 contributed to the biosynthesis of xanthotoxol. Notably, CYP71AZ19 originated from a proximal duplication event of CYP71AZ18, specific to A. dahurica , subsequently undergoing neofunctionalization. Accessible chromatin regions (ACRs), especially proximal ACRs, are correlated with higher gene expression levels, including the three validated genes involved in FC biosynthesis, showing potential to regulate metabolite biosynthesis. Our findings provide new insights into the biosynthetic pathway of FCs and the epigenetic regulation of metabolite biosynthesis. ### Competing Interest Statement The authors have declared no competing interest.

Amomi Fructus (Sharen, AF) is a traditional Chinese medicine (TCM) from three source species (or varieties), including Wurfbainia villosa var. villosa (WVV), W. villosa var. xanthioides (WVX), or W. longiligularis (WL). Among them, WVV has been transplanted from its top-geoherb region, Guangdong, to its current main production area, Yunnan, for >50 years in China. However, the genetic and transcriptomic differentiation among multiple AF source species (or varieties) and between the origin and transplanted populations of WVV is unknown. In our study, the observed overall higher expression of terpenoid biosynthesis genes in WVV than in WVX provided possible evidence for the better pharmacological effect of WVV. We also screened six candidate borneol dehydrogenases (BDHs) that potentially catalyzed borneol into camphor in WVV and functionally verified them. Highly expressed genes at the P2 stage of WVV, Wv05G1424 and Wv05G1438, were capable of catalyzing the formation of camphor from (+)-borneol, (-)-borneol and DL-isoborneol. Moreover, the BDH genes may experience independent evolution after acquiring the ancestral copies, and the following tandem duplications might account for the abundant camphor content in WVV. Furthermore, four populations of WVV, WVX, and WL are genetically differentiated, and the gene flow from WVX to WVV in Yunnan contributed to the greater genetic diversity in the introduced population (WVV-JH) than in its top-geoherb region (WVV-YC), which showed the lowest genetic diversity and might undergo genetic degradation. In addition, terpene synthesis (TPS) and BDH genes were selected among populations of multiple AF source species (or varieties) and between the top- and non-top-geoherb regions, which might explain the difference in metabolites between these populations. Our findings provide important guidance for the conservation, genetic improvement, and industrial development of the three source species (or varieties) and for identifying top-geoherbalism with molecular markers, and proper clinical application of AF.

Soybean (Glycine max L.) is a protein and oil crop grown worldwide. Its fitness may be reduced by deleterious mutations, whose identification and purging is desirable for crop breeding. In the published whole-genome re-sequenced data of 2214 soybean accessions, including 221 wild soybean, 1132 landrace cultivars and 861 improved soybean lines, we identified 115,275 deleterious single-nucleotide polymorphisms (SNPs). Numbers of deleterious alleles increased from wild soybeans to landraces and decreased from landraces to modern improved lines. Genes in selective-sweep regions showed fewer deleterious mutations than the remaining genes. Deleterious mutations explained 4.3%–48% more phenotypic variation than randomly selected SNPs for resistance to soybean cyst nematode race 2 (SCN2), soybean cyst nematode race 3 (SCN3) and soybean mosaic virus race 3 (SMV3). These findings illustrate how mutation load has shifted during soybean domestication, expansion and improvement and provide candidate sites for breeding out deleterious mutations in soybean by genome editing and/or conventional breeding focused on the selection of progeny with fewer deleterious alleles.

Future breeding is likely to involve the detection and removal of deleterious alleles, which are mutations that negatively affect crop fitness. However, little is known about the prevalence of such mutations and their effects on phenotypic traits in the context of modern crop breeding. To address this, we examined the number and frequency of deleterious mutations in 350 elite maize inbred lines developed over the past few decades in China and the United States. Our findings reveal an accumulation of weakly deleterious mutations and a decrease in strongly deleterious mutations, indicating the dominant effects of genetic drift and purifying selection for the two types of mutations, respectively. We also discovered that slightly deleterious mutations, when at lower frequencies, were more likely to be heterozygous in the developed hybrids. This is consistent with complementation as a potential explanation for heterosis. Subsequently, we found that deleterious mutations accounted for more of the variation in phenotypic traits than nondeleterious mutations with matched minor allele frequencies, especially for traits related to leaf angle and flowering time. Moreover, we detected fewer deleterious mutations in the promoter and gene body regions of differentially expressed genes across breeding eras than in nondifferentially expressed genes. Overall, our results provide a comprehensive assessment of the prevalence and impact of deleterious mutations in modern maize breeding and establish a useful baseline for future maize improvement efforts.

SUMMARYCoumarins are natural products with important medicinal values, and include simple coumarins, furanocoumarins and pyranocoumarins. Female ginseng (Angelica sinensis) is a renowned herb with abundant coumarins, originated in China and known for the treatment of female ailments for thousands of years. The molecular basis of simple coumarin biosynthesis in A. sinensis and the evolutionary history of the genes involved in furanocoumarin biosynthesis are largely unknown. Here, we generated the first chromosome‐scale genome of A. sinensis. It has a genome size of 2.37 Gb, which was generated by combining PacBio and Hi‐C sequencing technologies. The genome was predicted to contain 43 202 protein‐coding genes dispersed mainly on 11 pseudochromosomes. We not only provided evidence for whole‐genome duplication (WGD) specifically occurring in the Apioideae subfamily, but also demonstrated the vital role of tandem duplication for phenylpropanoid biosynthesis in A. sinensis. Combined analyses of transcriptomic and metabolomic data revealed key genes and candidate transcription factors regulating simple coumarin biosynthesis. Furthermore, phylogenomic synteny network analyses suggested prenyltransferase genes involved in furanocoumarin biosynthesis evolved independently in the Moraceae, Fabaceae, Rutaceae and Apiaceae after ζ and ε WGD. Our work sheds light on coumarin biosynthesis, and provides a benchmark for accelerating genetic research and molecular breeding in A. sinensis.

Verticillium wilt is a vascular disease causing tremendous damage to cotton production worldwide. However, our knowledge of the mechanisms of cotton resistance or susceptibility to this disease is very limited. In this study, we compared the defense transcriptomes of cotton (Gossypium hirsutum) cultivars Shidalukang 1 (Verticillium dahliae resistant, HR) and Junmian 1 (V. dahliae susceptible, HS) before and after V. dahliae infection, identified hub genes of the network associated with responses to V. dahliae infection, and functionally characterized one of the hub genes involved in biosynthesis of lignin and phenolics. We identified 6,831 differentially expressed genes (DEGs) between the basal transcriptomes of HR and HS; 3,685 and 3,239 of these DEGs were induced in HR and HS, respectively, at different time points after V. dahliae infection. KEGG pathway analysis indicated that DEGs were enriched for genes involved in lignin biosynthesis. In all, 23 hub genes were identified based on a weighted gene coexpression network analysis of the 6,831 DEGs and their expression profiles at different time points after V. dahliae infection. Knockdown of Gh4CL30, one of the hub genes related to the lignin biosynthesis pathway, by virus-induced gene silencing, led to a decreased content of flavonoids, lignin, and S monomer but an increased content of G monomer, G/S lignin monomer, caffeic acid, and ferulic acid, and enhanced cotton resistance to V. dahliae. These results suggest that Gh4CL30 is a key gene modulating the outputs of different branches of the lignin biosynthesis pathway, and provide new insights into cotton resistance to V. dahliae. [Formula: see text] Copyright © 2021 The Author(s). This is an open access article distributed under the CC BY-NC-ND 4.0 International license .

Verticillium wilt, caused by Verticillium dahliae , is one of the most damaging and widespread soil-borne cotton diseases. The molecular mechanisms underlying the cotton defense against V. dahliae remain largely elusive. Here, we compared the transcriptional differences between Upland cotton cultivars: one highly resistant (HR; Shidalukang 1) and one highly susceptible (HS; Junmian 1). This was done at multiple time points after V. dahliae inoculation, which identified 2010 and 1275 differentially expressed genes (DEGs) in HR and HS, respectively. Plant hormone signal transduction-related genes were enriched in HR, whereas genes related to lignin biosynthesis were enriched in both HR and HS. Weighted gene co-expression network analysis (WGCNA) using the 2868 non-redundant genes differentially expressed between the V. dahliae infected and uninfected samples in HR or HS identified 10 different gene network modules and 22 hub genes with a potential role in regulating cotton defense against V. dahliae infection. GhGDH2 , encoding glutamate dehydrogenase (GDH), was selected for functional characterization. Suppressing the expression level of GhGDH2 by virus-induced gene silencing (VIGS) in HS led to inhibition of the salicylic acid (SA) biosynthesis/signaling pathways and activation of the jasmonic acid (JA) biosynthesis/signaling pathways, which resulted in an increase of 42.1% JA content and a reduction of 78.9% SA content in cotton roots, and consequently enhanced V. dahliae resistance. Our finding provides new insights on the molecular mechanisms of cotton resistance to V. dahliae infection and candidate genes for breeding V. dahliae resistance cotton cultivars by genetic modification.