Search for: All records

Uncovering causal structures from observational data is crucial for understanding complex systems and making informed decisions. While reinforcement learning (RL) has shown promise in identifying these structures in the form of a directed acyclic graph (DAG), existing methods often lack efficiency, making them unsuitable for online applications. In this paper, we propose MARLIN, an efficient multi-agent RL-based approach for incremental DAG learning. MARLIN uses a DAG generation policy that maps a continuous real-valued space to the DAG space as an intra-batch strategy, then incorporates two RL agents—state-specific and state-invariant—to uncover causal relationships and integrates these agents into an incremental learning framework. Furthermore, the framework leverages a factored action space to enhance parallelization efficiency. Extensive experiments on synthetic and real datasets demonstrate that MARLIN outperforms state-of-the-art methods in terms of both efficiency and effectiveness. 

Abstract Nicotinamide adenine dinucleotide (NAD+)-capped RNAs (NAD-RNAs) are widely present in prokaryotic and eukaryotic transcriptomes. However, their regulation during development remains unclear. Here, we investigated NAD-RNA regulation during receptacle ripening in strawberry (Fragaria × ananassa, Fa), a model for nonclimacteric fruit ripening. We performed “click reaction”-based sequencing of NAD-RNAs (SPAAC-NAD-seq) to globally profile NAD-RNAs during ripening. This identified 9,109 NAD-RNAs produced from protein-coding genes (NAD-mRNAs), 6,479 from transposable elements (NAD-TE-RNAs), 4 from the mitochondrial genome, and 11 from the chloroplast genome. The number of NAD-RNAs tended to decrease, corresponding with lower abundance of the NAD+ cap during ripening. In addition, we characterized the function of the NAD-RNA scavenger Nudix (Nucleoside diphosphate linked moiety X)-Type Motif 19 (FaNUDT19). FaNUDT19 expression increased approximately 3.5-fold during ripening. Biochemically, FaNUDT19 could hydrolyze multiple 5′ caps, including NAD+, ATP, AMP, and m7G caps, with the strongest activity on NAD-RNAs. Indeed, FaNUDT19 overexpression in strawberry selectively affected a subset of NAD-mRNAs, allowing for the persistence of stable levels of ripening-related genes, such as PEROXIDASE (FaPOD). These results suggest that FaNUDT19 affects strawberry ripening by stabilizing mRNAs that produce proteins promoting coloration. Overall, our findings highlight the importance of dynamic NAD+ decapping by FaNUDT19 in strawberry ripening.