Autopolyploid crops play a central role in global agriculture, yet their complex genomes pose significant barriers to genetic improvement. High allelic div

Modern Hopfield networks are energy-based associative memory models whose performance critically depends on the structure and optimization of their energy functions. While recent formulations substantially improve storage capacity, the resulting non-convex energy landscapes are often optimized using heuristic update rules that can be sensitive to initialization and may not provide monotonic energy descent or rigorous convergence guarantees. In this work, we propose a new energy formulation for modern Hopfield networks together with a principled iterative optimization scheme. The proposed energy admits a natural decomposition that allows optimization via the concave–convex procedure (CCCP), yielding well-defined network dynamics with guaranteed energy descent beyond classical Hopfield updates. We establish fundamental theoretical properties of the proposed framework, including non-negativity, boundedness, and monotonic decrease in the energy along iterations. In particular, we prove that the induced dynamics converge to a stationary point of the energy function, providing explicit convergence guarantees for the resulting Hopfield-type model. We further evaluate the proposed approach on synthetic classification tasks and compare its optimization behavior with that of the original Hopfield network and several standard machine learning baselines. Experimental results demonstrate improved stability, convergence behavior, and competitive classification performance. We also validate the approach on real-world benchmark datasets to demonstrate utility beyond controlled experiments. Overall, this work provides a theoretically grounded energy-based optimization framework for modern Hopfield networks, clarifying the role of principled optimization in achieving stable and convergent associative memory dynamics.

Anthocyanin-based visual reporters enable rapid, non-destructive identification of transgenic tissues, but their pigment output may not be physiologically neutral during organogenesis, raising concerns about their suitability as selection markers in transformation pipelines. To address this, we examined the regeneration performance of constitutive ROSEA1 expression across four eudicot species from three families: tomato, petunia, begonia, and marigold. Stable ROSEA1-overexpressing lines were evaluated in tomato and petunia, whereas transformation-stage assays were performed in begonia and marigold. Regeneration frequencies were quantified relative to controls, and transcript-level changes were assessed for anthocyanin biosynthetic genes and regeneration regulators; complementation assays with PLT5 co-expression and hormonal manipulations were also performed. In tomato and petunia, ROSEA1-overexpressing lines exhibited reduced regeneration frequencies and elevated anthocyanin accumulation, with two independent petunia lines showing that stronger pigment activation corresponded to more severe regeneration defects. In tomato, ROSEA1 coordinately upregulated CHI, F3H, F3′5′H, and DFR and was associated with reduced expression of PLT5, WUS, and LBD16 during early regeneration. PLT5 co-expression partially restored regeneration with modestly reduced pigmentation, indicating that pigment output alone does not fully explain the phenotype. The penalty extended to begonia and marigold, with genotype- and hormone-dependent severity in marigold. These findings indicate a context-dependent regeneration penalty associated with constitutive ROSEA1 expression and suggest that hormone optimization, genotype selection, and developmental support may help mitigate this limitation in some contexts.

Leaf senescence directly affects lettuce quality and postharvest shelf life, but the regulatory roles of miR319-targeted and non-target CIN-TCP transcription factors remain unclear. Here, we examined whether the miR319–TCP module controls lettuce leaf senescence through separable genetic branches. MIR319 overexpression delayed dark-induced senescence, whereas STTM-mediated miR319 suppression accelerated chlorophyll loss, photosynthetic decline, and senescence-marker activation. Disruption of the miR319-targeted gene LsTCP4 phenocopied MIR319 overexpression, supporting LsTCP4 as a pro-senescence factor downstream of miR319. We further found that the miR319 non-target CIN gene LsTCP17 also promoted senescence, as tcp17 leaves retained more chlorophyll than wild type during dark treatment. Genetic combinations showed that tcp17 enhanced chlorophyll retention in the OX319 background and partially rescued the accelerated senescence phenotype of S319, indicating that LsTCP17 acts through a route separable from the miR319-targeted branch. Together, these results reveal a split CIN-TCP architecture in which miR319-targeted LsTCP4 and non-target LsTCP17 provide parallel pro-senescence inputs, offering a genetic framework for targeted improvement of lettuce quality.