{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"bioRxiv"}],"indexed":{"date-parts":[[2026,9,11]],"date-time":"2026-09-11T15:10:24Z","timestamp":1789139424808,"version":"build-2803163510"},"posted":{"date-parts":[[2026,9,8]]},"group-title":"Bioinformatics","reference-count":53,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,9,8]],"date-time":"2026-09-08T00:00:00Z","timestamp":1788825600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,9,8]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:p>\n                  Predicting biological function from genomic sequence remains a major challenge in computational and systems biology. Here, we tested whether the genomic language model Evo2, which encodes context-dependent DNA sequence patterns into embeddings, enables prediction of essential genes in fungi, a phenotype central to fungal biology and antifungal target discovery. We found that model performance was constrained not by the type or complexity of the downstream classifier, but by the biological information contained in Evo2 DNA embeddings. Specifically, the information recoverable from these embeddings progressively declined for biological features further downstream of DNA sequence, revealing a bottleneck for predicting higher-order cellular phenotypes. We alleviated this bottleneck by integrating Evo2 embeddings with two sequence-informed, system-level features: ortholog-based essentiality and protein-protein interactions. The multimodal framework demonstrated consistent performance both within and across three evolutionarily divergent yeasts (\n                  <jats:italic>Candida albicans<\/jats:italic>\n                  ,\n                  <jats:italic>Saccharomyces cerevisiae<\/jats:italic>\n                  , and\n                  <jats:italic>Schizosaccharomyces pombe<\/jats:italic>\n                  ), and its predictions were supported by published experimental evidence when transferred to the filamentous mold\n                  <jats:italic>Aspergillus fumigatus<\/jats:italic>\n                  . These results establish our framework as a transferrable tool for predicting essential genes across fungal genomes, including species with limited or no experimentally determined essentially data.\n                <\/jats:p>\n                <jats:sec>\n                  <jats:title>Importance<\/jats:title>\n                  <jats:p>\n                    Essential genes in fungi are promising targets for antifungal drug development, yet experimental, genome-wide essentiality screening remains slow, resource intensive, and difficult to scale. As a result, comprehensive gene essentiality profiles exist for only a few model species, including\n                    <jats:italic>Candida albicans<\/jats:italic>\n                    ,\n                    <jats:italic>Saccharomyces cerevisiae<\/jats:italic>\n                    , and\n                    <jats:italic>Schizosaccharomyces pombe<\/jats:italic>\n                    . In this work, we demonstrate that genomic language models can be leveraged through transfer learning to predict fungal essential genes directly from genomic sequence, either without training data from the target species or after fine-tuning on limited target species data. Our transfer learning approach offers a promising strategy to accelerate antifungal target discovery across understudied fungal pathogens and clinical isolates.\n                  <\/jats:p>\n                <\/jats:sec>","DOI":"10.64898\/2026.09.03.749190","type":"posted-content","created":{"date-parts":[[2026,9,9]],"date-time":"2026-09-09T03:55:16Z","timestamp":1788926116000},"source":"Crossref","is-referenced-by-count":0,"title":["A Transferable Genomic Language Model Framework for Fungal Gene Essentiality Prediction"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8474-1196","authenticated-orcid":false,"given":"Chen","family":"Liao","sequence":"first","affiliation":[{"name":"Department of Microbiology and Immunology, Geisel School of Medicine, Dartmouth, Hanover, New Hampshire, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hannah G.","family":"Thomas","sequence":"additional","affiliation":[{"name":"Department of Microbiology and Immunology, Geisel School of Medicine, Dartmouth, Hanover, New Hampshire, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026091107350723000_2026.09.03.749190v1.1","doi-asserted-by":"crossref","first-page":"101531","DOI":"10.1016\/j.cels.2026.101531","article-title":"Artificial intelligence for microbiology and microbiome research","volume":"17","year":"2026","journal-title":"Cell Syst"},{"key":"2026091107350723000_2026.09.03.749190v1.2","doi-asserted-by":"publisher","DOI":"10.1186\/1471-2105-16-S14-S1"},{"key":"2026091107350723000_2026.09.03.749190v1.3","doi-asserted-by":"publisher","DOI":"10.1093\/genetics\/iyae035"},{"key":"2026091107350723000_2026.09.03.749190v1.4","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pcbi.1011563"},{"key":"2026091107350723000_2026.09.03.749190v1.5","first-page":"1466","article-title":"Predicting natural variation in the yeast phenotypic landscape with machine learning","volume":"21","year":"2025","journal-title":"Mol Syst Biol"},{"key":"2026091107350723000_2026.09.03.749190v1.6","doi-asserted-by":"crossref","first-page":"aaiag005","DOI":"10.1093\/molecular-omics\/aaiag005","article-title":"Multi-omics integration in genome-scale metabolic models: a review of constraint-based approaches","volume":"22","year":"2026","journal-title":"Mol Omics"},{"key":"2026091107350723000_2026.09.03.749190v1.7","doi-asserted-by":"publisher","DOI":"10.1038\/nrg.2017.74"},{"key":"2026091107350723000_2026.09.03.749190v1.8","doi-asserted-by":"crossref","first-page":"101515","DOI":"10.1016\/j.cels.2025.101515","article-title":"The modifiers that cause changes in gene essentiality","volume":"17","year":"2026","journal-title":"Cell Syst"},{"key":"2026091107350723000_2026.09.03.749190v1.9","doi-asserted-by":"publisher","DOI":"10.1126\/science.1150021"},{"key":"2026091107350723000_2026.09.03.749190v1.10","doi-asserted-by":"publisher","DOI":"10.1126\/science.aaf1420"},{"key":"2026091107350723000_2026.09.03.749190v1.11","doi-asserted-by":"publisher","DOI":"10.1371\/journal.ppat.0030024"},{"key":"2026091107350723000_2026.09.03.749190v1.12","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1009845107"},{"key":"2026091107350723000_2026.09.03.749190v1.13","doi-asserted-by":"publisher","DOI":"10.1128\/mBio.02048-18"},{"key":"2026091107350723000_2026.09.03.749190v1.14","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-021-26850-3"},{"key":"2026091107350723000_2026.09.03.749190v1.15","doi-asserted-by":"crossref","first-page":"e3003184","DOI":"10.1371\/journal.pbio.3003184","article-title":"Landscape of essential growth and fluconazole-resistance genes in the human fungal pathogen Cryptococcus neoformans","volume":"23","year":"2025","journal-title":"PLOS Biol"},{"key":"2026091107350723000_2026.09.03.749190v1.16","doi-asserted-by":"publisher","DOI":"10.1101\/gr.5144106"},{"key":"2026091107350723000_2026.09.03.749190v1.17","doi-asserted-by":"crossref","first-page":"785","DOI":"10.1016\/j.csbj.2019.05.008","article-title":"An Evaluation of Machine Learning Approaches for the Prediction of Essential Genes in Eukaryotes Using Protein Sequence-Derived Features","volume":"17","year":"2019","journal-title":"Comput Struct Biotechnol J"},{"key":"2026091107350723000_2026.09.03.749190v1.18","doi-asserted-by":"publisher","DOI":"10.1186\/s12859-017-1884-5"},{"key":"2026091107350723000_2026.09.03.749190v1.19","doi-asserted-by":"crossref","first-page":"265","DOI":"10.1186\/1471-2164-7-265","article-title":"Towards the identification of essential genes using targeted genome sequencing and comparative analysis","volume":"7","year":"2006","journal-title":"BMC Genom"},{"key":"2026091107350723000_2026.09.03.749190v1.20","doi-asserted-by":"publisher","DOI":"10.15252\/msb.202110427"},{"key":"2026091107350723000_2026.09.03.749190v1.21","doi-asserted-by":"crossref","first-page":"910","DOI":"10.1186\/1471-2164-14-910","article-title":"A new computational strategy for predicting essential genes","volume":"14","year":"2013","journal-title":"BMC Genom"},{"key":"2026091107350723000_2026.09.03.749190v1.22","doi-asserted-by":"crossref","first-page":"lqab110","DOI":"10.1093\/nargab\/lqab110","article-title":"Identifying essential genes across eukaryotes by machine learning","volume":"3","year":"2021","journal-title":"NAR Genom Bioinform"},{"key":"2026091107350723000_2026.09.03.749190v1.23","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiolchem.2014.01.011"},{"key":"2026091107350723000_2026.09.03.749190v1.24","doi-asserted-by":"crossref","unstructured":"Schonfeld E , Vendrow E , Vendrow J , Schonfeld E . 2021. On the relation of gene essentiality to intron structure: a computational and deep learning approach. Life Sci Alliance 4.","DOI":"10.26508\/lsa.202000951"},{"key":"2026091107350723000_2026.09.03.749190v1.25","doi-asserted-by":"publisher","DOI":"10.1016\/j.tig.2024.11.013"},{"key":"2026091107350723000_2026.09.03.749190v1.26","doi-asserted-by":"crossref","first-page":"346","DOI":"10.1038\/s42256-025-01007-9","article-title":"Transformers and genome language models","volume":"7","year":"2025","journal-title":"Nat Mach Intell"},{"key":"2026091107350723000_2026.09.03.749190v1.27","doi-asserted-by":"crossref","first-page":"1349","DOI":"10.1038\/s41586-026-10176-5","article-title":"Genome modelling and design across all domains of life with Evo 2","volume":"652","year":"2026","journal-title":"Nature"},{"key":"2026091107350723000_2026.09.03.749190v1.28","doi-asserted-by":"publisher","DOI":"10.1016\/j.mib.2024.102521"},{"key":"2026091107350723000_2026.09.03.749190v1.29","doi-asserted-by":"publisher","DOI":"10.1128\/EC.2.2.247-255.2003"},{"key":"2026091107350723000_2026.09.03.749190v1.30","doi-asserted-by":"publisher","DOI":"10.1016\/S1087-1845(03)00090-2"},{"key":"2026091107350723000_2026.09.03.749190v1.31","doi-asserted-by":"crossref","first-page":"1077","DOI":"10.1093\/genetics\/161.3.1077","article-title":"Characterization of essential genes by parasexual genetics in the human fungal pathogen Aspergillus fumigatus: impact of genomic rearrangements associated with electroporation of DNA","volume":"161","year":"2002","journal-title":"Genetics"},{"key":"2026091107350723000_2026.09.03.749190v1.32","first-page":"196","article-title":"Comprehensive prediction and analysis of human protein essentiality based on a pretrained large language model","volume":"5","year":"2025","journal-title":"Nat Comput Sci"},{"key":"2026091107350723000_2026.09.03.749190v1.33","first-page":"bbad472","article-title":"\u2018Bingo\u2019\u2014a large language model- and graph neural network-based workflow for the prediction of essential genes from protein data","volume":"25","year":"2024","journal-title":"Brief Bioinform"},{"key":"2026091107350723000_2026.09.03.749190v1.34","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pcbi.1000140"},{"key":"2026091107350723000_2026.09.03.749190v1.35","doi-asserted-by":"crossref","first-page":"109449","DOI":"10.1016\/j.compbiomed.2024.109449","article-title":"Protein-protein interaction detection using deep learning: A survey, comparative analysis, and experimental evaluation","volume":"185","year":"2025","journal-title":"Comput Biol Med"},{"key":"2026091107350723000_2026.09.03.749190v1.36","doi-asserted-by":"crossref","first-page":"e2610619123","DOI":"10.1073\/pnas.2610619123","article-title":"Linear-time prediction of proteome-scale microbial protein interactions","volume":"123","year":"2026","journal-title":"Proc Natl Acad Sci U S A"},{"key":"2026091107350723000_2026.09.03.749190v1.37","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkae1113"},{"key":"2026091107350723000_2026.09.03.749190v1.38","doi-asserted-by":"crossref","first-page":"1327","DOI":"10.1038\/s41592-026-03126-6","article-title":"OrthoFinder: improved phylogenetic orthology inference with enhanced accuracy and scalability","volume":"23","year":"2026","journal-title":"Nat Methods"},{"key":"2026091107350723000_2026.09.03.749190v1.39","doi-asserted-by":"crossref","unstructured":"Grover A , Leskovec J . 2016. node2vec: Scalable Feature Learning for Networks, p. 855\u2013864. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Association for Computing Machinery, New York, NY, USA.","DOI":"10.1145\/2939672.2939754"},{"key":"2026091107350723000_2026.09.03.749190v1.40","doi-asserted-by":"crossref","unstructured":"Tsai Y-HH , Bai S , Liang PP , Kolter JZ , Morency L-P , Salakhutdinov R . 2019. Multimodal Transformer for Unaligned Multimodal Language Sequences, p. 6558\u20136569. In Korhonen, A , Traum, D , M\u00e0rquez, L (eds.), Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. Association for Computational Linguistics, Florence, Italy.","DOI":"10.18653\/v1\/P19-1656"},{"key":"2026091107350723000_2026.09.03.749190v1.41","unstructured":"Sundararajan M , Taly A , Yan Q . 2017. Axiomatic Attribution for Deep Networks, p. 3319\u20133328. In Proceedings of the 34th International Conference on Machine Learning. PMLR."},{"key":"2026091107350723000_2026.09.03.749190v1.42","unstructured":"Shrikumar A , Tian K , Avsec \u017d , Shcherbina A , Banerjee A , Sharmin M , Nair S , Kundaje A . 2018. Technical Note on Transcription Factor Motif Discovery from Importance Scores (TF-MoDISco) version 0.5.6.5. arXiv.org. https:\/\/arxiv.org\/abs\/1811.00416v5. Retrieved 2 August 2026."},{"key":"2026091107350723000_2026.09.03.749190v1.43","doi-asserted-by":"publisher","DOI":"10.1128\/EC.00324-09"},{"key":"2026091107350723000_2026.09.03.749190v1.44","doi-asserted-by":"crossref","unstructured":"Yan K , Stanley M , Kowalski B , Raimi OG , Ferenbach AT , Wei P , Fang W , Aalten DMF van . 2022. Genetic validation of Aspergillus fumigatus phosphoglucomutase as a viable therapeutic target in invasive aspergillosis. J Biol Chem 298.","DOI":"10.1016\/j.jbc.2022.102003"},{"key":"2026091107350723000_2026.09.03.749190v1.45","doi-asserted-by":"publisher","DOI":"10.1042\/BSR20130053"},{"key":"2026091107350723000_2026.09.03.749190v1.46","doi-asserted-by":"crossref","first-page":"245","DOI":"10.1111\/mmi.14706","article-title":"Genetic and structural validation of phosphomannomutase as a cell wall target in Aspergillus fumigatus","volume":"116","year":"2021","journal-title":"Mol Microbiol"},{"key":"2026091107350723000_2026.09.03.749190v1.47","doi-asserted-by":"publisher","DOI":"10.1111\/mmi.12290"},{"key":"2026091107350723000_2026.09.03.749190v1.48","doi-asserted-by":"publisher","DOI":"10.1128\/EC.00107-10"},{"key":"2026091107350723000_2026.09.03.749190v1.49","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.0601456103"},{"key":"2026091107350723000_2026.09.03.749190v1.50","doi-asserted-by":"crossref","first-page":"e3003409","DOI":"10.1371\/journal.pbio.3003409","article-title":"Expansion of the functional genomics GRACE library reveals genes relevant for temperature-dependent fitness in Candida albicans","volume":"23","year":"2025","journal-title":"PLOS Biol"},{"key":"2026091107350723000_2026.09.03.749190v1.51","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2009.07896"},{"key":"2026091107350723000_2026.09.03.749190v1.52","doi-asserted-by":"publisher","DOI":"10.1093\/NAR\/GKAD347"},{"key":"2026091107350723000_2026.09.03.749190v1.53","doi-asserted-by":"publisher","DOI":"10.1186\/1471-2105-10-421"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.09.03.749190","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,11]],"date-time":"2026-09-11T14:35:17Z","timestamp":1789137317000},"score":1,"resource":{"primary":{"URL":"http:\/\/biorxiv.org\/lookup\/doi\/10.64898\/2026.09.03.749190"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,8]]},"references-count":53,"URL":"https:\/\/doi.org\/10.64898\/2026.09.03.749190","relation":{},"subject":[],"published":{"date-parts":[[2026,9,8]]},"subtype":"preprint"}}