{
  "id": 12464184,
  "title": "sORF-Trans2MS: a two-module deep learning framework for sORF translation and MS-supported microprotein prediction",
  "url": "https://urgent.news/2026/10/06/sorf-trans2ms-a-two-module-deep-learning-framework-for-sorf",
  "topic": "science",
  "section": "Science",
  "published": "2026-10-06T00:00:00.000Z",
  "source": {
    "name": "bioRxiv",
    "slug": "biorxiv",
    "url": "https://www.biorxiv.org/content/10.64898/2026.09.29.755517v1?rss=1"
  },
  "original_language": "en",
  "account": "Ribosome profiling is commonly employed to identify translated small open reading frames (sORFs); however, existing prediction models demonstrate limited performance on novel sORF datasets. Notably, numerous microproteins with robust mass spectrometry (MS) evidence lack corresponding support in public Ribo-seq resources, indicating that Ribo-seq-supported sORF detection and MS-supported microprotein prediction may be distinct yet complementary tasks. To address these challenges, the researchers introduced sORF-Trans2MS, a two-module deep learning framework for predicting Ribo-supported sORFs and MS-supported microproteins. The first module employs RNA-FM representations of upstream, ORF, and downstream regions with a Transformer encoder, designed for sORF translation prediction. The second module integrates pretrained RNA-FM representations and ESM-2 protein representations using multi-scale CNN encoders, aimed at predicting microproteins supported by MS but not Ribo-seq data. The researchers trained Module I on 9,579 Ribo-supported sORFs and matched background sORFs, while Module II utilized 14,697 MS-supported (MS++Ribo-) microproteins and matched backgrounds. Module I demonstrated superior internal performance with an AUROC of 0.870 and an AUROC of 0.930 on an independent GENCODE dataset. Ablation analysis revealed that upstream RNA context, particularly a distinct signal around 50 nt upstream of AUG, played a crucial role. Module II achieved an AUROC/AUPR of 0.804/0.797 on an independent mass spectrometry dataset. Upon model interpretation, the researchers identified upstream RNA context and the protein N terminus as significant factors, while literature-supported microproteins further illustrated the complementary roles of the two modules. In summary, sORF-Trans2MS enhances the prediction and generalization of Ribo-supported sORFs and expands computational discovery to MS-supported microproteins not represented in current public Ribo-seq resources.",
  "summary": "Ribosome profiling is widely used to identify translated small open reading frames (sORFs), but existing prediction models often generalize poorly to newly collected sORF datasets. In addition, many microproteins with strong mass spectrometry (MS) evidence lack support in public Ribo-seq resources, suggesting that Ribo-supported sORF and MS-supported microprotein detection may represent…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}