A deep learning and large language hybrid workflow for omics interpretation
- 8 January 2026
- journal article
- research article
- Published by Springer Nature in Nature Biomedical Engineering
- Vol. 10 (8) , 1-22
Abstract
Profiling molecular panorama from massive omics data identifies regulatory networks in cells but requires mechanistic interpretation and experimental follow up. Here we combine deep learning and large language model reasoning to develop a hybrid workflow for omics interpretation, called LyMOI. LyMOI incorporates GPT-3.5 for biological knowledge reasoning and a large graph model with graph convolutional networks (GCNs). The large graph model integrates evolutionarily conserved protein interactions and uses hierarchical fine-tuning to predict context-specific molecular regulators from multi-omics data. GPT-3.5 then generates machine chain-of-thought (CoT) to mechanistically interpret their roles in biological systems. Focusing on autophagy, LyMOI mechanistically interprets 1.3 TB transcriptomic, proteomic and phosphoproteomic data and expands the knowledge of autophagy regulators. We also show that LyMOI highlights two human oncoproteins, CTSL and FAM98A, for enhancing autophagy upon treatment with disulfiram (DSF), an antitumour agent. Silencing these genes in vitro attenuates DSF-mediated autophagy and suppresses cancer cell proliferation. Strikingly, DSF treatment with Z-FY-CHO, a CTSL-specific inhibitor previously used for preventing SARS-CoV-2 infection, potently inhibits tumour growth in vivo.Keywords
This publication has 81 references indexed in Scilit:
- The acquisition of resistance to TNFα in breast cancer cells is associated with constitutive activation of autophagy as revealed by a transcriptome analysis using a custom microarrayAutophagy, 2011
- Inhibition of cathepsin L lowers the apoptotic threshold of glioblastoma cells by up-regulating p53 and transcription of caspases 3 and 7Apoptosis, 2011
- The Autophagy Database: an all-inclusive information resource on autophagy that provides nourishment for researchNucleic Acids Research, 2010
- DeathBase: a database on structure, evolution and function of proteins involved in apoptosis and other forms of cell deathCell Death & Differentiation, 2010
- Deciphering Protein Kinase Specificity Through Large-Scale Analysis of Yeast Phosphorylation Site MotifsScience Signaling, 2010
- iRefIndex: A consolidated protein interaction database with provenanceBMC Bioinformatics, 2008
- Loss of p53 induces epidermal growth factor receptor promoter activity in normal human keratinocytesOncogene, 2008
- Direct Phosphorylation and Activation of a Nim1-related Kinase Gin4 by Elm1 in Budding YeastJournal of Biological Chemistry, 2006
- Integrative analysis of the cancer transcriptomeNature Genetics, 2005
- Annotation Transfer Between Genomes: Protein–Protein Interologs and Protein–DNA RegulogsGenome Research, 2004