Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2511.21124

Moonshine.jl: a Julia package for genome-scale model-based ancestral recombination graph inference

Patrick Fournier·Fabrice Larribe
View original source ↗Read source document ↗Export metadata ↓

Abstract

The ancestral recombination graph (ARG) is the model of choice in statistical genetics to model population ancestries. Software capable of simulating ARGs on a genome scale within a reasonable amount of time are now widely available for most practical use cases. While the inverse problem of inferring ancestries from a sample of haplotypes has seen major progress in the last decade, it does not enjoy the same level of advancement as its counterpart. Up until recently, even moderately sized samples could only be handled using heuristics. In recent years, the possibility of model-based inference for datasets closer to "real world" scenarios has become a reality, largely due to the development of threading-based samplers. This article introduces Moonshine.jl, a Julia package that has the ability, among other things, to infer ARGs for samples of thousands of human haplotypes of sizes on the order of hundreds of megabases within a reasonable amount of time. On recent hardware, our package is able to infer an ARG for samples of densely haplotyped (over one marker/kilobase) human chromosomes of sizes up to 10000 in well under a day on data simulated by msprime. Scaling up simulation on a compute cluster is straightforward thanks to a strictly single-threaded implementation. While model-based, it does not resort to threading but rather places restrictions on probability distributions typically used in simulation software in order to enforce sample consistency. In addition to being efficient, a strong emphasis is placed on ease of use and integration into the biostatistical software ecosystem.

Explore related subjects

q-bio.GN

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Patrick Fournier, Fabrice Larribe. 2026-03-17. Moonshine.jl: a Julia package for genome-scale model-based ancestral recombination graph inference. https://doi.org/10.3389/fgene.2026.1753780

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Transcriptomic Models for Immunotherapy Response Prediction Show Limited Cross-cohort Generalisability

Immune checkpoint inhibitors (ICIs) have transformed cancer therapy; yet substantial proportion of patients exhibit intrinsic or acquired resistance, making accurate pre-treatment response prediction a critical unmet need. Transcriptomics-based biomarkers derived from bulk and single-cell RNA sequencing (scRNA-seq) offer a promising avenue for capturing tumour-immune interactions, yet the cross-cohort generalisability of existing prediction models remains unclear.We systematically benchmark nine state-of-the-art transcriptomic ICI response predictors, five bulk RNA-seq-based models (COMPASS, IRNet, NetBio, IKCScore, and TNBC-ICI) and four scRNA-seq-based models (PRECISE, DeepGeneX, Tres and scCURE), using publicly available independent datasets unseen during model development. Overall, predictive performance was modest: bulk RNA-seq models performed at or near chance level across most cohorts, while scRNA-seq models showed only marginal improvements. Pathway-level analyses revealed sparse and inconsistent biomarker signals across models. Although scRNA-seq-based predictors converged on immune-related programs such as allograft rejection, bulk RNA-seq-based models exhibited little reproducible overlap. PRECISE and NetBio identified the most coherent immune-related themes, whereas IRNet predominantly captured metabolic pathways weakly aligned with ICI biology. Together, these findings demonstrate the limited cross-cohort robustness and biological consistency of current transcriptomic ICI prediction models, underscoring the need for improved domain adaptation, standardised preprocessing, and biologically grounded model design.

q-bio.GN↗
arXiv2026

Large Language Model Agents for Evidence Based Genetic Disease Severity Classification

Disease severity classification for genetic conditions is subjective and labor-intensive, creating bottlenecks in genomic screening, where commercial panels vary widely in size and overlap. We developed an autonomous AI agent integrating Reasoning and Acting (ReAct) with Retrieval-Augmented Generation (RAG) to classify 10,211 Human Phenotype Ontology terms. It uses American College of Medical Genetics (ACMG)-endorsed severity guidelines and American College of Obstetricians and Gynecologists (ACOG) quality-of-life criteria to retrieve PubMed literature, generate interpretable reasoning chains, and independently verify claims. At the phenotype level, using expert-curated cohorts, the agent achieved 93.55% accuracy (MCC 0.9237) with 82.6% to 91.4% of claims supported by direct evidence or valid inferences. Gene-level severity was aggregated across 8,738 pairs, identifying 3,283 autosomal recessive pairs with severe or profound presentations. External validation showed 95.2% concordance with Mackenzie's Mission gene list. This system enables standardized panel design by providing reliable, automated classification supported by direct evidence.

q-bio.GN↗
arXiv2026

Harmonised benchmarking of foundation models for single-cell and spatial transcriptomics reveals context-dependent generalisation

Single-cell and spatial foundation models promise transferable biological representations, yet their generality remains largely untested across modalities, biological domains and analytical tasks. We benchmarked six representative models, Nicheformer, CellPLM, scGPT-spatial, GenePT, scELMo and Novae, using a harmonised framework spanning scRNA-seq, spatial transcriptomics and Perturb-seq. We evaluated zero-shot and continually pretrained clustering, supervised annotation, marker-gene concordance and perturbation prediction. Model performance was strongly conditional: expression-trained cell-level transformers best resolved many cell-identity tasks, spatial and graph-aware models better preserved tissue architecture, and language-derived gene embeddings were competitive for selected perturbation-response metrics. No model dominated across tasks, and rankings shifted with modality, preprocessing, tokenisation, biological prior, domain shift and metric choice. This benchmark provides practical guidance for model selection and argues that future models should be judged by biological generalisation, interpretability and perturbation-grounded validity, not by scale or leaderboard performance alone.

q-bio.GN↗