Large language model (LLM) agents have recently demonstrated impressive capabilities in various domains like open-ended conversation and multi-step decision-making. However, it remains challenging for these agents to solve strategic language games, such as Werewolf, which demand both strategic decision-making and free-form language interactions. Existing LLM agents often suffer from intrinsic bias in their action distributions and limited exploration of the unbounded text action space, resulting in suboptimal performance. To address these challenges, we propose Latent Space Policy Optimization (LSPO), an iterative framework that combines game-theoretic methods with LLM fine-tuning to build strategic language agents. LSPO leverages the observation that while the language space is combinatorially large, the underlying strategy space is relatively compact. We first map free-form utterances into a finite latent strategy space, yielding an abstracted extensive-form game. Then we apply game-theoretic methods like Counterfactual Regret Minimization (CFR) to optimize the policy in the latent space. Finally, we fine-tune the LLM via Direct Preference Optimization (DPO) to align with the learned policy. By iteratively alternating between these steps, our LSPO agents progressively enhance both strategic reasoning and language communication. Experiment on the Werewolf game shows that our agents iteratively expand the strategy space with improving performance and outperform existing Werewolf agents, underscoring their effectiveness in free-form language games with strategic interactions.

Relatedness within genomic cohorts is a potential source of bias for many genetic analyses. Existing tools to break relatedness are phenotype-naïve; they indiscriminately remove subjects to break relationship, risking the loss of valuable data, especially in studies targeting uncommon and rare phenotypes. To address this limitation, we developed the Kinship Decouple and Phenotype Selection (KDPS) tool, with a novel algorithm designed to enhance the precision of subject selection in genetic and epidemiological research by incorporating phenotype prioritization. KDPS separates related individuals by considering relatedness (kinship or identity by descent) scores and allows prioritizing subjects based on phenotypes of interest. This approach enables the retention of valuable subjects for analysis, even in the face of necessary exclusions due to relatedness. Furthermore, KDPS accommodates a wide range of phenotypes, including quantitative and categorical, and allows for customization to either prioritize specific phenotypes or maximization of sample size. In simulations based on the UK Biobank dataset and real-world datasets, KDPS demonstrated significant improvements in retention of subjects with prioritized phenotypes and computational efficiency compared to previously published software. The ability of this method to process biobank-scale studies within practical timeframes marks the ability of this method to process biobank-scale studies within practical timeframes and marks a considerable advancement over existing techniques. KDPS offers tailored solutions for complex analytical challenges and broad applicability in genetics and epidemiology research. To our knowledge, KDPS is the first tool to perform phenotype-aware decoupling, paving the way for more powerful genetic and epidemiological analyses.

SUMMARY: Currently, most computational methods estimate the expression of circular RNAs (circRNAs) using the number of sequencing reads that support back-splicing junctions (BSJ) in RNA-seq data, which may introduce biased estimation of circRNA expression due to the uneven distribution of sequencing reads. To overcome this, we previously developed a model-based strategy for circRNA quantification, enabling consideration of sequencing reads from the entire transcript. Yet, the lack of exact transcript structure of circRNAs may limit its accuracy. Here, we proposed a substantially improved circRNA quantification tool, AQUARIUM, by introducing the full-length RNA structure of circular isoforms. We assessed its performance in circRNA quantification using both biological and simulated rRNA-depleted RNA-seq datasets, and demonstrated its superior performance at both BSJ and isoform level. AVAILABILITY AND IMPLEMENTATION: AQUARIUM is freely available at https://github.com/wanjun-group-seu/AQUARIUM. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: Circular RNAs (circRNAs) are a class of non-coding RNAs that are widely expressed in various cell lines and tissues of many organisms. Although the exact function of many circRNAs is largely unknown, the cell type-and tissue-specific circRNA expression has implicated their crucial functions in many biological processes. Hence, the quantification of circRNA expression from high-throughput RNA-seq data is becoming important to ascertain. Although many model-based methods have been developed to quantify linear RNA expression from RNA-seq data, these methods are not applicable to circRNA quantification. RESULTS: Here, we proposed a novel strategy that transforms circular transcripts to pseudo-linear transcripts and estimates the expression values of both circular and linear transcripts using an existing model-based algorithm, Sailfish. The new strategy can accurately estimate transcript expression of both linear and circular transcripts from RNA-seq data. Several factors, such as gene length, amount of expression and the ratio of circular to linear transcripts, had impacts on quantification performance of circular transcripts. In comparison to count-based tools, the new computational framework had superior performance in estimating the amount of circRNA expression from both simulated and real ribosomal RNA-depleted (rRNA-depleted) RNA-seq datasets. On the other hand, the consideration of circular transcripts in expression quantification from rRNA-depleted RNA-seq data showed substantial increased accuracy of linear transcript expression. Our proposed strategy was implemented in a program named Sailfish-cir. AVAILABILITY AND IMPLEMENTATION: Sailfish-cir is freely available at https://github.com/zerodel/Sailfish-cir . CONTACT: [email protected] or [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

SUMMARY: Phylogenetics, likelihood, evolution and complexity (PLEX) is a flexible and fast Bayesian Markov chain Monte Carlo software program for large-scale analysis of nucleotide and amino acid data using complex evolutionary models in a phylogenetic framework. The program gains large speed improvements over standard approaches by implementing 'partial sampling of substitution histories', a data augmentation approach that can reduce data analysis times from months to minutes on large comparative datasets. A variety of nucleotide and amino acid substitution models are currently implemented, including non-reversible and site-heterogeneous mixture models. Due to efficient algorithms that scale well with data size and model complexity, PLEX can be used to make inferences from hundreds to thousands of taxa in only minutes on a desktop computer. It also performs probabilistic ancestral sequence reconstruction. Future versions will support detection of co-evolutionary interactions between sites, probabilistic tests of convergent evolution and rigorous testing of evolutionary hypotheses in a Bayesian framework. AVAILABILITY AND IMPLEMENTATION: PLEX v1.0 is licensed under GPL. Source code and documentation will be available for download at www.evolutionarygenomics.com/ProgramsData/PLEX. PLEX is implemented in C++ and supported on Linux, Mac OS X and other platforms supporting standard C++ compilers. Example data, control files, documentation and accessory Perl scripts are available from the website. CONTACT: [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Recent studies have suggested that the thermodynamic stability of mRNA secondary structure near the start codon can regulate translation efficiency in Escherichia coli, and that translation is more efficient the less stable the secondary structure. We survey the complete genomes of 340 species for signals of reduced mRNA secondary structure near the start codon. Our analysis includes bacteria, archaea, fungi, plants, insects, fishes, birds, and mammals. We find that nearly all species show evidence for reduced mRNA stability near the start codon. The reduction in stability generally increases with increasing genomic GC content. In prokaryotes, the reduction also increases with decreasing optimal growth temperature. Within genomes, there is variation in the stability among genes, and this variation correlates with gene GC content, codon bias, and gene expression level. For birds and mammals, however, we do not find a genome-wide trend of reduced mRNA stability near the start codon. Yet the most GC rich genes in these organisms do show such a signal. We conclude that reduced stability of the mRNA secondary structure near the start codon is a universal feature of all cellular life. We suggest that the origin of this reduction is selection for efficient recognition of the start codon by initiator-tRNA.