Research Appraisals
Evidence-based critical appraisals of the latest medical research, systematically evaluated using Oxford CEBM methodology.
Showing 11 appraisals
Journal of medical Internet research
Initial-Visit Specialty Triage in Rare Diseases Using Large Language Models: Retrospective Benchmarking Study
BACKGROUND: Specialty triage at first contact is an overlooked step in early diagnostic pathways for rare diseases. Patients often present with overlapping, multisystem, and atypical manifestations, making first-visit specialty selection challenging and potentially prolonging diagnostic pathways. OBJECTIVE: The aim of this study is to evaluate the accuracy, response time, and consistency of large language models (LLMs) for initial-visit specialty triage in rare diseases across multiple datasets, and to compare their performance with registered nurses and nonmedical participants. METHODS: In this retrospective benchmarking study, we used 5 rare disease datasets: a publication-derived case set, 3 RareBench-derived datasets, and a Facial phenotype-Gene-Disease Dataset-derived set. Fourteen LLMs were evaluated over 5 independent runs per case. Performance was assessed using accuracy, response time, and consistency, with subgroup analyses by model accessibility, reasoning mode, parameter scale, and phenotype count. Human comparison was conducted on the publication-derived case set using registered nurses and nonmedical participants. RESULTS: Across datasets, model accuracy ranged from 0.4378 to 0.7141. Claude-opus-4-5 achieved the highest accuracy (0.7141) and consistency (0.9653), averaging 10.79 seconds per case. GPT-5.1 had the shortest response time (3.39 s/case) and high accuracy (0.6948). Proprietary models had numerically higher average accuracy than open-weight models (0.6973 vs 0.6365). Nonthinking models achieved higher average accuracy than thinking models (0.6789 vs 0.5826) and had shorter response times, although this exploratory comparison was based on a small number of thinking models. Accuracy varied by phenotype count, with higher performance in cases with 1 to 2 or more than 14 phenotypes. On the publication-derived case set, LLMs achieved higher average accuracy than registered nurses and nonmedical participants (0.5978 vs 0.4914 and 0.4573). CONCLUSIONS: LLMs showed potential as assistive tools for initial-visit specialty triage in rare diseases. Model choice, reasoning mode, and phenotype information density influenced performance, but subgroup findings should be interpreted cautiously. Future work should evaluate LLM-based specialty triage in prospective clinical settings and develop clinician-supervised workflows with traceable evidence support.
24 July 2026
Read appraisal →Orphanet journal of rare diseases
Clinical outcomes in alpha-mannosidosis: a systematic review of therapeutic approaches
BACKGROUND: Alpha mannosidosis (AM) is a rare lysosomal storage disorder caused by a deficiency in the α-mannosidase enzyme, resulting in impaired glycoprotein metabolism within lysosomes. Enzyme dysfunction is attributed to an autosomal recessive mutation in the MAN2B1 gene. Affected individuals present with a broad spectrum of manifestations, including developmental delays, cognitive decline, musculoskeletal abnormalities, hearing difficulties, and recurrent infections. Current therapeutic options are limited to hematopoietic stem cell transplantation and the more recently developed enzyme replacement therapy. OBJECTIVE: The aim of this review was to evaluate and compare the therapeutic outcomes, benefits and challenges associated with Hematopoietic stem cell transplantation (HSCT) and enzyme replacement therapy (ERT) in the treatment of AM. METHODS: A systematic search across PubMed, MEDLINE, EMBASE, the Cochrane Library, OMIM, and ScienceDirect identified 12 original studies from 307 records. The data are presented narratively due to the scarcity of literature and the heterogeneity of study designs and interventions. RESULTS: A total of 28 patients who received hematopoietic stem cell transplantation showed improvements in preserving neurocognitive function and skeletal stabilization and reduced infection rates, especially when performed at relatively young ages. However, this treatment carries significant risks, including infections, graft-versus-host disease, and increased morbidity and mortality, particularly in older patients. Conversely, enzyme replacement therapy was administered to 75 patients, who demonstrated a favorable safety profile, enhanced respiratory function, reduced skeletal abnormalities, and improved overall quality of life. However, enzyme replacement therapy has limited efficacy in preventing neurocognitive decline and requires lifelong administration. CONCLUSION: Both interventions yield better outcomes when initiated early, particularly before cognitive deterioration becomes significant. This review emphasizes the importance of a timely diagnosis to optimize treatment outcomes and prevent severe complications.
23 July 2026
Read appraisal →Nucleic acids research
FABIAN-variant 2026: improved prediction of the effects of DNA variants on transcription factor binding
Variants in promoters and enhancers can alter the binding of transcription factors (TFs), but their functional assessment remains difficult. FABIAN-variant is a web application that predicts the effects of DNA variants on TF binding by comparing position weight matrix (PWM) and transcription factor flexible model (TFFM) scores between reference and variant alleles. Here, we present FABIAN-variant 2026, a major update that expands the prediction model library from ~5000 to over 40 000 models for >1500 human TFs, sourced from nine PWM databases and including 1290 TFFMs. The application now supports the mouse genome (GRCm38 and GRCm39) with over 35 000 models for >1100 mouse TFs. An optional BPNet deep learning scorer provides neural network-based binding predictions for 240 human TFs. Known TF binding site information has been expanded from three to five sources. Predictions for over 1400 heterodimer TF complexes have been added. The web server has been rewritten in Rust and the scoring engine optimized, reducing runtime by ~70%. A RESTful JSON API and a standalone command-line version enable programmatic access and local high-throughput analysis. FABIAN-variant 2026 is available at https://fabianapp.org/variant26/. The web server is free and open to all users and there is no login requirement.
13 July 2026
Read appraisal →Nucleic acids research
SNPnexus: an enhanced web platform for large-scale and multi-sample variant analysis (2025 update)
SNPnexus is a long-standing web-based platform for the functional annotation and prioritization of genetic variants. Since its previous release, SNPnexus has undergone substantial backend re-engineering resulting in major performance improvements and a complete restructuring of the underlying datasets. A redesigned interface now enables larger queries and multi-sample analysis enabling comparative workflows such as identifying shared pathogenic variants in disease cohorts and divergent mutations in cancer evolution studies. SNPnexus increased its capacity to 150 000 variants per query and introduced pre-annotation filters that allow users to restrict analyses to selected genes or genomic regions, improving efficiency while enabling targeted interrogation of high-priority targets. SNPnexus integrates updated and expanded annotations across both GRCh37 and GRCh38, covering genomic consequences, in silico pathogenicity predictions, population allele frequencies, evolutionary conservation, regulatory elements, biological pathways, and clinical associations. The refreshed result interface provides interactive visualizations for single-sample and cohort-level outputs, together with advanced filtering and export options. Optional user accounts now support query history and real-time job monitoring while preserving full, unregistered access for all users. SNPnexus remains free and open to all users without login requirements at https://snpnexus.org/.
13 July 2026
Read appraisal →Briefings in bioinformatics
Literature-informed gene extraction and ranking for multimodal data fusion.
Published biomedical experiments provide an increasingly rich collection of results and identify genes potentially involved in diverse biological mechanisms. However, individual studies are often confined to narrow experimental contexts and are restricted to single omics layers. Cross-study knowledge aggregation can broaden this perspective and enable the construction of global, context-aware gene rankings. Recent developments in natural language processing have made large-scale literature mining increasingly feasible. This enables the systematic extraction and fusion of symbolic knowledge from published experiments. We present pathXcite, a software that extracts genes associated with specific contexts, such as diseases or biological mechanisms from the literature, and ranks them by contextual relevance. These relevance-based gene rankings can compress a scientific context into a symbolic representation. This representation enables diverse downstream analyses, including cross-context comparisons, network-based analysis, enrichment analysis, and integration with experimental omics data. In multiple use cases, we show how our extraction and fusion strategy can be applied to uncover hidden aspects in biological data.
6 July 2026
Read appraisal →Briefings in bioinformatics
Bridging local-global transmembrane protein contexts with contrastive pretraining for alignment-free pathogenicity prediction.
Predicting the pathogenic consequences of protein mutations is a cornerstone of precision medicine, yet it remains a formidable challenge for transmembrane proteins (TMPs), a clinically vital class of drug targets. Existing computational methods are often hampered by their reliance on evolutionary data and fail to model TMP-specific biophysical constraints. Here, we introduce Memo-Patho, a deep learning framework for robust, alignment-free pathogenicity prediction of TMP variants. The core innovation is a within-protein, label-informed supervised contrastive pretraining strategy that learns sequence-encoded biophysical signatures distinguishing pathogenic and benign variants by directly comparing them within the same protein context. By fusing sequence-level representations from protein language models with local structural proxies derived from sequence, Memo-Patho achieves accurate predictions without multiple sequence alignments or experimental structures. Across diverse TMP benchmarks and under protein-level group splits, Memo-Patho consistently outperforms leading predictors, achieving up to 0.93 accuracy, and it transfers to an independent KCNQ1 ion-channel cohort without re-training. Its resource-efficient, alignment-free design enables routine large-scale screening when evolutionary or structural data are sparse. Conceptually, Memo-Patho addresses a key gap by directly learning discriminative, sequence-anchored signatures pertinent to TMP-specific constraints, offering a principled and generalizable foundation for research-use clinical variant triage and proteome-wide mutation-effect modeling.
6 July 2026
Read appraisal →Journal of the American Medical Informatics Association : JAMIA
Disparate language and model effects on AI-based translation and recognition of genetic conditions.
INTRODUCTION: Artificial intelligence (AI) is increasingly prevalent. Patients and clinicians may use AI-based tools in many different languages. OBJECTIVE: To investigate AI translation tools for descriptions of genetic conditions and how AI identification of genetic conditions is affected by translations. MATERIALS AND METHODS: We used Neural machine translation (NMT) and large language-model (LLM) translation to translate descriptions of 40 genetic conditions into 191 and 93 languages, respectively. Excluding translations retaining English medical terms verbatim, we respectively focused on 139 and 70 languages. After assessing translations, we assessed the ability of 3 proprietary and 3 open-weight general LLMs to identify conditions in the translations. We analyzed how accuracy was affected by the conditions' prevalence in the literature, and attributes of the languages (the script, language family, and prevalence of the language in training sources). We also investigated adaptive translation for select languages. RESULTS: We found significant differences in condition identification based on the translation method, condition, language, and prediction model. The accuracy of some models was more affected than others by factors like the conditions' literature prevalence, language script, family, and language prevalence. Adaptive translation for select languages did not improve translations or diagnostic accuracy with the 3 tested LLMs. However, further analysis with 1 language showed that this approach was more effective with smaller LLMs. CONCLUSIONS: AI-based translation has variable performance, which can affect the ability of AI models to recognize genetic conditions. These findings should inform safe medical AI use to support consistent performance in different languages.
2 July 2026
Read appraisal →Journal of pediatric surgery
Artificial intelligence in rare pediatric solid tumor research and clinical care: A scoping review
BACKGROUND: Clinical and research advances for children with solid tumors are limited by their rare nature. Artificial intelligence (AI) and machine learning (ML) hold potential for advancing diagnosis, risk stratification, and treatment in rare diseases where individual patients contribute high-dimensional data. This review characterizes current AI/ML applications in rare pediatric solid tumors. METHODS: PubMed, Embase, and Web-of-Science Core Collection databases were queried for relevant articles published before February 10, 2025. Eligible studies applied AI/ML methods to study rare tumors in pediatric populations (≤19 years). Articles were evaluated for tumor diagnosis, AI/ML methods, clinical applications, comparators, effectiveness, and interpretability. RESULTS: Twenty-three studies (2009-2025) were included. Hepatoblastoma (12/23) and pediatric thyroid cancers (4/23) were most frequently studied. Supervised learning predominated (20/23), followed by unsupervised (9/23), deep learning (4/23), and hybrid/ensemble models (6/23). Applications included diagnosis (12/23), prognosis (10/23), and risk stratification (9/23). Twenty studies reported effectiveness measures, with many models achieving AUCs >0.85. In comparative analysis (17/23), AI/ML often equaled or exceeded expert consensus, traditional models, or alternative algorithms. Eight studies reported external validation. CONCLUSIONS: Current AI/ML research in rare pediatric extra-cranial solid tumors focuses on diagnosis, risk stratification, and prognosis, often outperforming traditional methods. Future work should prioritize external validation and clinical applicability.
1 July 2026
Read appraisal →Human genetics
AI in variant analysis: fast track to genetic diagnoses
While falling costs have expanded access to genomic sequencing, clinical utility is frequently hindered by the challenge of interpreting complex genetic data. Variant analysis for rare disease patients especially requires significant time and expertise, creating a bottleneck that delays diagnostics. Although advances in genetic variant classification have improved diagnostic precision, they have also increased the identification of variants of uncertain significance (VUSs), widening the interpretation gap between data generation and clinical actionability. The high prevalence of VUSs can lead to false reassurance or psychological distress by misinterpretting inconclusive results. We propose that artificial intelligence (AI) is a critical clinical decision-support tool for bridging this gap, offering a scalable framework to optimize variant interpretation and shorten the diagnostic odyssey. While reclassification ultimately requires biological evidence that AI cannot replace, these tools serve as essential aggregators and prioritizers, especially as guidelines transition toward the upcoming quantitative ACMG v4 framework. We advocate integrating AI throughout the genetic diagnostic workflow-from initial phenotyping to variant prioritization-to facilitate data-driven, personalized treatment. We outline current AI-assisted approaches and discuss anticipated challenges in this pursuit, such as privacy, training data bias and quality, model explainability, and the necessity of a total product life cycle for validation. To address these challenges, we provide recommendations for "human-in-the-loop" design and intuitive workflow integration to ensure AI tools meet the highest standards of precision, reproducibility, and transparency to maximize adoption. By standardizing AI across the variant analysis pipeline, we can fast-track the path to genetic diagnoses, effectively bridging the interpretation gap and enabling rapid delivery of personalized medical interventions.
28 June 2026
Read appraisal →Nucleic acids research
SpliceSelectNet: a hierarchical Transformer-based deep learning model for splice site prediction
Accurate RNA splicing is essential for gene expression and protein function, yet the mechanisms governing splice site recognition remain incompletely understood. Aberrant splicing caused by mutations can lead to severe diseases, including cancer and genetic disorders, underscoring the need for accurate computational tools to predict splice sites and detect disruptions. Existing methods have made significant advances in splice site prediction but are often limited in handling long-range dependencies due to high computational costs, a factor critical to splicing regulation. Moreover, many models lack interpretability, hindering efforts to elucidate the underlying biological mechanisms. Here, we present SpliceSelectNet (SSNet), a hierarchical Transformer-based deep learning model that predicts splice sites from DNA sequences spanning up to 100 kb. By integrating local and global attention mechanisms, SSNet efficiently captures both proximal and distal regulatory signals while maintaining single-nucleotide resolution. Across multiple benchmark datasets, SSNet achieves state-of-the-art performance in splice site prediction and aberrant splicing detection. Systematic in silico mutagenesis demonstrates that attention scores reflect functional sequence importance, supporting their biological relevance. Long-range sequence perturbation experiments further show that SSNet captures distal regulatory effects beyond conventional receptive fields. Together, these results establish SSNet as a biologically interpretable framework for modeling long-range splicing regulation from genomic sequence.
23 June 2026
Read appraisal →Signal transduction and targeted therapy
Liposomal doxorubicin plus nab-paclitaxel with/without chemoradiotherapy in head and neck adenoid cystic carcinoma: single-arm phase II study
Given no standard therapies, this single-arm, open-label, multicenter phase II study was conducted in unresectable locally advanced or recurrent/metastatic adenoid cystic carcinoma of the head and neck (LA or R/M ACCHN). Patients received chemotherapeutic nanomedicines (20 mg/m2 liposomal doxorubicin plus 120 mg/m2 nab-paclitaxel on days 1 and 8) for three 21-day cycles. Subsequently, patients eligible for radiotherapy received concurrent chemoradiotherapy (cCRT; helical tomotherapy or intensity-modulated radiotherapy and nab-paclitaxel), whereas those ineligible for radiotherapy continued 2-3 cycles of original chemotherapy. Between May 27, 2021, and November 13, 2024, 31 enrolled patients were evaluable for efficacy and safety. The primary endpoint objective response rate (ORR) was 90.3% (95% CI, 74.2-98.0%), with 15 (48.4%) complete responses and 13 (41.9%) partial responses, meeting the prespecified primary endpoint. The median duration of response was 18.5 (95% CI, 0-45.8) months. With a median follow-up of 15.7 months by data cutoff (June 25, 2025), the median progression-free survival (PFS) was 25.7 (95% CI, 6.5-44.8) months; the 1-year PFS was 77.1%. The median overall survival (OS) was not reached (NR), with a 1-year OS of 90.0%. Patients who received cCRT had a prolonged PFS (NR vs. 12.8 months) and OS (NR vs. 21.2 months) compared to those who did not. Grade 3 treatment-related adverse events (TRAEs) occurred in 8 (25.8%) patients. No ≥grade 4 TRAEs or treatment-related deaths occurred. This study demonstrated tumor responses among the highest reported to date, with a manageable safety profile in unresectable LA or R/M ACCHN. Chemotherapeutic nanomedicines, especially sequential cCRT for appropriate patients, may represent an effective and safe option.
14 June 2026
Read appraisal →