Explore the vast range of titles available.

The accurate identification of soybean growth stages is critical for optimizing agricultural interventions, where mistimed treatments can result in yield losses ranging from 2.5% to 40%. Existing deep learning approaches remain limited in scope, targeting isolated developmental phases rather than providing comprehensive phenological coverage. This paper presents a novel object detection architecture DELTA-SoyStage, combining an EfficientNet backbone with a lightweight ChannelMapper neck and a newly proposed DELTA (Denoising Enhanced Lightweight Task Alignment) detection head for soybean growth stage classification. We introduce a dataset of 17,204 labeled RGB images spanning nine growth stages from emergence (VE) through full maturity (R8), collected under controlled greenhouse conditions with diverse imaging angles and lighting variations. DELTA-SoyStage achieves 73.9% average precision with only 24.4 GFLOPs computational cost, demonstrating 4.2× fewer FLOPs than the best-performing baseline (DINO-Swin: 74.7% AP, 102.5 GFLOPs) with only 0.8% accuracy difference. The lightweight DELTA head combined with the efficient ChannelMapper neck requires only 8.3 M parameters-a 43.5% reduction compared to standard architectures-while maintaining competitive accuracy. Extensive ablation studies validate key design choices including task alignment mechanisms, multi-scale feature extraction strategies, and encoder-decoder depth configurations. The proposed model's computational efficiency makes it suitable for deployment on resource-constrained edge devices in precision agriculture applications, enabling timely decision-making without reliance on cloud infrastructure.

Two types of resistant soybean ( Glycine max (L.) Merr.) sources are widely used against soybean cyst nematode (SCN, Heterodera glycines Ichinohe). These include Peking-type soybean, whose resistance requires both the rhg1-a and Rhg4 alleles, and PI 88788-type soybean, whose resistance requires only the rhg1-b allele. Multiple copy number of PI 88788-type GmSNAP18 , GmAAT , and GmWI12 in one genomic segment simultaneously contribute to rhg1-b resistance. Using an integrated set of genetic and genomic approaches, we demonstrate that the rhg1-a Peking-type GmSNAP18 is sufficient for resistance to SCN in combination with Rhg4 . The two SNAPs (soluble NSF attachment proteins) differ by only five amino acids. Our findings suggest that Peking-type GmSNAP18 is performing a different role in SCN resistance than PI 88788-type GmSNAP18 . As such, this is an example of a pathogen resistance gene that has evolved to underlie two types of resistance, yet ensure the same function within a single plant species. Peking-type and PI 88788-type soybeans are both used as sources of resistance against soybean cyst nematode. Here the authors show that in contrast to PI 88788-type resistance, where GmSNAP18 acts in combination with two neighbouring genes, in Peking-type resistance GmSNAP18 acts with GmSHMT08 to confer resistance.

Soybean (Glycine max (L.) Merr.) is a globally important legume crop valued as a major source of plant-based protein and edible oil. Understanding the transcriptional programs underlying tissue-specific development is essential for improving seed quality and agronomic performance. Here, we present an integrative transcriptomic analysis of soybean based on 12 samples representing key seed developmental stages—including globular, heart, cotyledon, embryo, dry seed, mid-mature, and late-mature—and vegetative and reproductive tissues, including leaf, root, stem, flower bud, and seedling at 6 days after imbibition (6 DAI). Following data preprocessing and filtering, 54,880 genes were retained for downstream analysis. Principal component analysis revealed clear separation between seed and non-seed tissues, indicating that tissue identity is the dominant driver of transcriptomic variation. Analysis of the top 100 most variable genes further highlighted distinct expression modules associated with seed maturation and vegetative growth. Differential expression analysis identified 9785 genes exhibiting significant expression differences between seed and non-seed tissues, including 1139 upregulated and 8646 downregulated genes under relaxed statistical thresholds. Functional characterization of seed-upregulated genes revealed enrichment of biological processes related to storage metabolism, embryo development, and stress protection mechanisms associated with desiccation tolerance. In addition, co-expression network and correlation analyses demonstrated strong transcriptional coherence among seed tissues and distinct clustering of vegetative organs. Together, these results provide a comprehensive systems-level overview of transcriptional organization across soybean tissues and identify candidate gene sets relevant to seed biology, functional genomics, and crop improvement.

Key messageSoybean acyl-ACP thioesterase gene family have been characterized; GmFATA1A mutants were discovered to confer high oleic acid, while GmFATB mutants presented low palmitic and high oleic acid seed content.Soybean oil stability and quality are primarily determined by the relative proportions of saturated versus unsaturated fatty acids. Commodity soybean typically contains 11% palmitic acid, as the primary saturated fatty acids. Reducing palmitic acid content is the principal approach to minimize the levels of saturated fatty acids in soybean. Though high palmitic acid enhances oxidative stability of soybean oil, it is negatively correlated with oil and oleic acid content and can cause coronary heart diseases for humans. For plants, acyl–acyl carrier protein (ACP) thioesterases (TEs) are a group of enzymes to hydrolyze acyl group and release free fatty acid from plastid. Among them, GmFATB1A has become the main target to genetically reduce the palmitic acid content in soybean. However, the role of members in soybean acyl-ACP thioesterase gene family is largely unknown. In this study, we characterized two classes of TEs, GmFATA, and GmFATB in soybean. We also denominated two GmFATA members and discovered six additional members that belong to GmFATB gene family through phylogenetic, syntenic, and in silico analysis. Using TILLING-by-Sequencing+, we identified an allelic series of mutations in five soybean acyl-ACP thioesterase genes, including GmFATA1A, GmFATB1A, GmFATB1B, GmFATB2A, and GmFATB2B. Additionally, we discovered mutations at GmFATA1A to confer high oleic acid (up to 34.5%) content, while mutations at GmFATB presented low palmitic acid (as low as 5.6%) and high oleic acid (up to 36.5%) phenotypes. The obtained soybean mutants with altered fatty acid content can be used in soybean breeding program for improving soybean oil composition traits.

Nitrogen (N), phosphorus (P), and sulfur (S) are essential nutrients for plant health. Deficiencies in N, P, or S in plants lead to lower seed production and seed quality in grain crops, including soybean seed. Soybean seed is a source of protein, oil, essential amino acids, and minerals. These nutrients are essential for plant health, and maintaining N, P, and S levels in soybean seed is crucial for higher seed nutritional value and amino acids quality. There is limited information on genomic regions, candidate genes, and molecular markers associated with soybean seed N, P, and S. Two field experiments were carried out in two locations using a ‘Forrest’ × ‘Williams 82’ recombinant inbred lines (RIL) population. A 306 RIL population and 2075 SNP markers were used to create the genetic map. The results showed a wide range of N, P, and S concentrations in both locations among RIL population lines. Based on the broad-sense heritability (H 2 ), 91.7% of seed N concentration variation was due to genetic effects, followed by 48.2% for S seed concentration, and a heritability of close to zero for seed P concentration. Eleven QTL were identified for seed N, seven QTL for seed P, and nine QTL for seed S in two locations. All these QTL had a significant linkage to the trait as their LOD ranged from 2.5 to 6.48 in 2018 and from 2.75 to 128.72 in 2020. Two QTL for seed N ( qN-02 -[IL-2020] on Chr 4, and qN-03 -[IL-2020] on Chr 4 were identified at the marker Gm04_4687302-Gm04_7672403 and Gm04_7672403, and their LOD were 45.06 and 96.98, and their contribution to the phenotypic variation were 45.85% and 48.37%, respectively. The low heritability of P indicated a major interactions between the trait (P) and environment. Except for the seed N, P, and S QTL, identified on Chr 16, 11 QTL reported here were not previously identified and therefore are novel. Several functional genes encoding N-, P-, and S-proteins, enzymes, and transporters were identified and located within the QTL interval. To our knowledge, the QTL identified here on Chr 2 and 6 are novel and were not previously identified. Therefore, QTL, genes, and molecular markers discovered in this research will provide breeders with new knowledge and tools for soybean selection for optimum seed mineral nutritional qualities. Also, this new findings advance our knowledge of physiology and genetics of seed N, S, and P candidate genes for genetic engineering application.

Summary Soybean cyst nematode (SCN) is the most devastating plant‐parasitic nematode. Most commercial soybean varieties with SCN resistance are derived from PI88788. Resistance derived from PI88788 is breaking down due to narrow genetic background and SCN population shift. PI88788 requires mainly the rhg1‐b locus, while ‘Peking’ requires rhg1‐a and Rhg4 for SCN resistance. In the present study, whole genome re‐sequencing of 106 soybean lines was used to define the Rhg haplotypes and investigate their responses to the SCN HG‐Types. The analysis showed a comprehensive profile of SNPs and copy number variations (CNV) at these loci. CNV of rhg1 (GmSNAP18) only contributed towards resistance in lines derived from PI88788 and ‘Cloud’. At least 5.6 copies of the PI88788‐type rhg1 were required to confer SCN resistance, regardless of the Rhg4 (GmSHMT08) haplotype. However, when the GmSNAP18 copies dropped below 5.6, a ‘Peking’‐type GmSHMT08 haplotype was required to ensure SCN resistance. This points to a novel mechanism of epistasis between GmSNAP18 and GmSHMT08 involving minimum requirements for copy number. The presence of more Rhg4 copies confers resistance to multiple SCN races. Moreover, transcript abundance of the GmSHMT08 in root tissue correlates with more copies of the Rhg4 locus, reinforcing SCN resistance. Finally, haplotype analysis of the GmSHMT08 and GmSNAP18 promoters inferred additional levels of the resistance mechanism. This is the first report revealing the genetic basis of broad‐based resistance to SCN and providing new insight into epistasis, haplotype‐compatibility, CNV, promoter variation and its impact on broad‐based disease resistance in plants.

Phytoparasitic nematodes that are able to infect and reproduce on plants that are considered resistant are referred to as virulent. The mechanism(s) that virulent nematodes employ to evade or suppress host plant defenses are not well understood. Here we report the use of a genetic strategy (allelic imbalance analysis) to associate single nucleotide polymorphisms (SNPs) with nematode virulence genes in Heterodera glycines, the soybean cyst nematode (SCN). To accomplish this analysis, a custom SCN SNP array was developed and used to genotype SCN F3-derived populations grown on resistant and susceptible soybean plants. Three SNPs reproducibly showed allele imbalances between nematodes grown on resistant and susceptible plants. Two candidate SCN virulence genes that were tightly linked to the SNPs were identified. One SCN gene encoded biotin synthase (HgBioB), and the other encoded a bacterial-like protein containing a putative SNARE domain (HgSLP-1). The two genes mapped to two different linkage groups. HgBioB contained sequence polymorphisms between avirulent and virulent nematodes. However, the gene encoding HgSLP-1 had reduced copy number in virulent nematode populations and appears to produce multiple forms of the protein via intron retention and alternative splicing. We show that HgSLP-1 is an esophageal-gland protein that is secreted by the nematode during plant parasitism. Furthermore, in bacterial co-expression experiments, HgSLP-1 co-purified with the SCN resistance protein Rhg1 α-SNAP, suggesting that these two proteins physically interact. Collectively our data suggest that multiple SCN genes are involved in SCN virulence, and that HgSLP-1 may function as an avirulence protein and when absent it helps SCN evade host defenses.

In soybeans, eighteen members constitute the serine hydroxymethyltransferase ( GmSHMT ) gene family, of which the cytosolic-targeted GmSHMT08 c member has been reported to mediate resistance to soybean cyst nematode (SCN). This work presents a comprehensive study of the SHMT gene family members, including synteny, phylogeny, subcellular localizations, haplotypes, protein homology modeling, mutational, and expression analyses. Phylogenetic analysis showed that SHMT genes are divided into four classes reflecting their subcellular distribution (cytosol, nucleus, mitochondrion, and chloroplast). Subcellular localization of selected GmSHMT members supports their in-silico predictions and phylogenetic distribution. Expression and functional analyses showed that GmSHMT genes display many overlapping, but some divergent responses during SCN infection. Furthermore, mutational analysis reveals that all isolated EMS mutants that lose their resistance to SCN carry missense and nonsense mutations at the GmSHMT08c , but none of the Gmshmt08c mutants carried mutations in the other GmSHMT genes. Haplotype clustering analysis using the whole genome resequencing data from a collection of 106 diverse soybean germplams (15X) was performed to identify allelic variants and haplotypes within the GmSHMT gene family. Interestingly, only the cytosolic-localized GmSHMT08c presented SNP clusters that were associated with SCN resistance, supporting our mutational analysis. Although eight GmSHMT members respond to the nematode infestation, functional and mutational analysis has shown the absence of functional redundancy in resistance to SCN. Structural analysis and protein homology modeling showed the presence of spontaneous mutations at important residues within the GmSHMT proteins, suggesting the presence of altered enzyme activities based on substrate affinities. Due to the accumulation of mutations during the evolution of the soybean genome, the other GmSHMT members have undergone neofunctionalization and subfunctionalization events.

Soybean seed quality is influenced by its soluble sugar composition, with high sucrose content being desirable for nutritional and industrial applications. In contrast, excessive raffinose and stachyose levels are considered undesirable due to their adverse effects on gastrointestinal function in humans and monogastric animals. Therefore, developing soybean mutant lines with elevated sucrose content and optimal raffinose and stachyose content is desirable. In this study, we characterized twelve sucrose synthase genes through a comprehensive phylogenetic tree analysis, synteny analysis, gene structure evaluation, and variations in conserved domains. Additionally, we conducted a TILLING by Sequencing approach to identify EMS mutations in the characterized Sucrose synthase genes. Numerous mutations have been identified in soybean sucrose synthases that resulted in high sucrose content, including the sucrose synthases mutants SL446 (R582W) and F1115 (G249E) on Glyma.02G240400 with a sucrose content of 9.5% and 9.1%, respectively. The obtained soybean mutants with enhanced sugar content can be useful in breeding programs to improve soybean nutritional quality without potential developmental trade-offs.

The CLAVATA pathway that regulates stem cell numbers of the shoot apical meristem has exclusively been studied in Arabidopsis; as such insight into other species is warranted. In this study, a GmCLV1A mutant ( F-S562L ) with altered lateral organ development, and two mutants of GmNARK , isolated from a Forrest M2 population (EMS-mutated soybean) were studied. GmCLV1A and GmNARK encode for LRR receptor kinases, and share 92% of protein sequence. While GmNARK is critical for systemic regulation of nodulation (new organ made on the root through symbiosis), we show that GmCLV1A functions locally and has no apparent function in nodulation or root development. However, a recessive, loss-of-function mutation ( S562L ) in a putative S-glycosylation site of GmCLV1A causes stem nodal identity alterations as well as flower and pod abnormalities (deformed flower and pod). The mutant also exhibits a homeotic phenotype, displaying abnormal leaf development/number, vein-derived leaf emergence, and a thick, faciated stem. The mutant phenotype is also temperature-sensitive. Interestingly, a novel truncated version of GmCLV1A was identified upstream of GmCLV1A that is absent from GmNARK , but is present upstream of the GmNARK orthologues, MtSUNN and PvNARK . Taken together, our findings indicate that GmCLV1A acts on shoot architecture, whereas GmNARK , functions in controlling nodule numbers.