{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,12]],"date-time":"2026-09-12T09:06:38Z","timestamp":1789203998764,"version":"build-2803163510"},"reference-count":31,"publisher":"Springer Science and Business Media LLC","license":[{"start":{"date-parts":[[2026,9,12]],"date-time":"2026-09-12T00:00:00Z","timestamp":1789171200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,9,12]],"date-time":"2026-09-12T00:00:00Z","timestamp":1789171200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"name":"Centro Andaluz de Biolog\u00eda del Desarrollo"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["J Mol Evol"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>The Ras protein superfamily comprises small GTPases that share a conserved G-domain but differ in flanking regions, regulation, and cellular roles. Because existing classifications rely mainly on G-domain phylogeny, this superfamily provides a useful test case for assessing whether protein language model embeddings recover biologically meaningful sequence organization consistent with established evolutionary classifications. Here, we analyzed a curated Ras superfamily dataset using three classification schemes: the classical five-family view, a G-domain phylogeny-based classification, and UniProtKB family annotation classification. We compared embeddings from multiple protein language models using supervised classification, unsupervised clustering, and residue-level ablation. Sequence-derived embeddings recovered known Ras superfamily organization across analyses. In supervised analyses, simple linear classifiers achieved high performance, indicating that Ras family and subfamily information is linearly accessible from sequence-derived embeddings. In unsupervised analyses, ESM-C-300 layer 12 gave the strongest full-protein recovery of the G-domain phylogeny-based classification, whereas ProstT5 performed best for G-domain embeddings. Residue ablation identified recurrent candidate subfamily-informative regions both within and outside the G-domain, including signals mapping to structurally coherent regions associated with subfamily-specific regulatory or interaction-related functions. Together, these results indicate that protein language model embeddings provide an effective alignment-free representation of Ras family and subfamily organization, and can highlight candidate sequence regions associated with functional specialization from sequence alone.<\/jats:p>","DOI":"10.1007\/s00239-026-10349-w","type":"journal-article","created":{"date-parts":[[2026,9,12]],"date-time":"2026-09-12T08:11:13Z","timestamp":1789200673000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Protein Language Model Embeddings Recover Evolutionary and Functional Organization Across the Ras Superfamily"],"prefix":"10.1007","author":[{"given":"Ildefonso","family":"Cases","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alfonso","family":"Valencia","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ana M.","family":"Rojas","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,9,12]]},"reference":[{"key":"10349_CR1","doi-asserted-by":"publisher","first-page":"39","DOI":"10.1038\/nrm3255","volume":"13","author":"IM Ahearn","year":"2012","unstructured":"Ahearn IM, Haigis K, Bar-Sagi D, Philips MR (2012) Regulating the regulator: post-translational modification of RAS. Nat Rev Mol Cell Biol 13:39\u201351. https:\/\/doi.org\/10.1038\/nrm3255","journal-title":"Nat Rev Mol Cell Biol"},{"key":"10349_CR2","doi-asserted-by":"publisher","unstructured":"Alsamkary H, Elshaffei M, Elkerdawy M, Elnaggar A (2025) Ankh3: multi-task pretraining with sequence denoising and completion enhances protein representations. https:\/\/doi.org\/10.48550\/arXiv.2505.20052. http:\/\/arxiv.org\/abs\/2505.20052","DOI":"10.48550\/arXiv.2505.20052"},{"key":"10349_CR3","doi-asserted-by":"publisher","unstructured":"Angarola B, Ferguson SM (2020) Coordination of Rheb lysosomal membrane interactions with mTORC1 activation. https:\/\/doi.org\/10.12688\/f1000research.22367.1. https:\/\/f1000research.com\/articles\/9-450","DOI":"10.12688\/f1000research.22367.1"},{"key":"10349_CR4","doi-asserted-by":"publisher","first-page":"lqae078","DOI":"10.1093\/nargab\/lqae078","volume":"6","author":"I Barrios-N\u00fa\u00f1ez","year":"2024","unstructured":"Barrios-N\u00fa\u00f1ez I, Mart\u00ednez-Redondo GI, Medina-Burgos P et al (2024) Decoding functional proteome information in model organisms using protein language models. NAR Genom Bioinform 6:lqae078. https:\/\/doi.org\/10.1093\/nargab\/lqae078","journal-title":"NAR Genom Bioinform"},{"key":"10349_CR5","doi-asserted-by":"publisher","first-page":"3453","DOI":"10.1016\/j.yexcr.2008.09.005","volume":"314","author":"J Berthold","year":"2008","unstructured":"Berthold J, Schenkov\u00e1 K, Ramos S et al (2008) Characterization of RhoBTB-dependent Cul3 ubiquitin ligase complexes \u2014 Evidence for an autoregulatory mechanism. Exp Cell Res 314:3453\u20133465. https:\/\/doi.org\/10.1016\/j.yexcr.2008.09.005","journal-title":"Exp Cell Res"},{"key":"10349_CR6","doi-asserted-by":"publisher","first-page":"7112","DOI":"10.1109\/TPAMI.2021.3095381","volume":"44","author":"A Elnaggar","year":"2022","unstructured":"Elnaggar A, Heinzinger M, Dallago C et al (2022) ProtTrans: Toward Understanding the Language of Life Through Self-Supervised Learning. IEEE Trans Pattern Anal Mach Intell 44:7112\u20137127. https:\/\/doi.org\/10.1109\/TPAMI.2021.3095381","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"10349_CR7","doi-asserted-by":"publisher","first-page":"19794","DOI":"10.1074\/jbc.271.33.19794","volume":"271","author":"JL Freeman","year":"1996","unstructured":"Freeman JL, Abo A, Lambeth JD (1996) Rac Insert Region Is a Novel Effector Region That Is Implicated in the Activation of NADPH Oxidase, but Not PAK65 *. J Biol Chem 271:19794\u201319801. https:\/\/doi.org\/10.1074\/jbc.271.33.19794","journal-title":"J Biol Chem"},{"key":"10349_CR8","doi-asserted-by":"publisher","first-page":"4296","DOI":"10.1016\/j.febslet.2006.06.086","volume":"580","author":"D Gizachew","year":"2006","unstructured":"Gizachew D, Oswald R (2006) NMR structural studies of the myristoylated N-terminus of ADP ribosylation factor 6 (Arf6). FEBS Lett 580:4296\u20134301. https:\/\/doi.org\/10.1016\/j.febslet.2006.06.086","journal-title":"FEBS Lett"},{"key":"10349_CR9","first-page":"1","volume-title":"Ras Signaling: Methods and Protocols","author":"L Goitre","year":"2014","unstructured":"Goitre L, Trapani E, Trabalzini L, Retta SF (2014) The Ras superfamily of Small GTPases: The Unlocked Secrets. In: Trabalzini L, Retta SF (eds) Ras Signaling: Methods and Protocols. Humana, Totowa, NJ, pp 1\u201318"},{"key":"10349_CR10","doi-asserted-by":"publisher","first-page":"102986","DOI":"10.1016\/j.sbi.2025.102986","volume":"91","author":"M Heinzinger","year":"2025","unstructured":"Heinzinger M, Rost B (2025) Teaching AI to speak protein. Curr Opin Struct Biol 91:102986. https:\/\/doi.org\/10.1016\/j.sbi.2025.102986","journal-title":"Curr Opin Struct Biol"},{"key":"10349_CR11","doi-asserted-by":"publisher","first-page":"lqae150","DOI":"10.1093\/nargab\/lqae150","volume":"6","author":"M Heinzinger","year":"2024","unstructured":"Heinzinger M, Weissenow K, Sanchez JG et al (2024) Bilingual language model for protein sequence and structure. NAR Genomics Bioinf 6:lqae150. https:\/\/doi.org\/10.1093\/nargab\/lqae150","journal-title":"NAR Genomics Bioinf"},{"key":"10349_CR12","doi-asserted-by":"publisher","first-page":"1494","DOI":"10.1242\/jcs.02958","volume":"119","author":"I Hofmann","year":"2006","unstructured":"Hofmann I, Munro S (2006) An N-terminally acetylated Arf-like GTPase is localised to lysosomes and affects their motility. J Cell Sci 119:1494\u20131503. https:\/\/doi.org\/10.1242\/jcs.02958","journal-title":"J Cell Sci"},{"key":"10349_CR13","doi-asserted-by":"publisher","first-page":"13566","DOI":"10.1038\/s41598-024-64211-4","volume":"14","author":"LR Jahn","year":"2024","unstructured":"Jahn LR, Marquet C, Heinzinger M, Rost B (2024) Protein embeddings predict binding residues in disordered regions. Sci Rep 14:13566. https:\/\/doi.org\/10.1038\/s41598-024-64211-4","journal-title":"Sci Rep"},{"key":"10349_CR14","doi-asserted-by":"publisher","first-page":"5764","DOI":"10.1074\/jbc.M111.336487","volume":"287","author":"JL Johnson","year":"2012","unstructured":"Johnson JL, Erickson JW, Cerione RA (2012) C-terminal Di-arginine Motif of Cdc42 Protein Is Essential for Binding to Phosphatidylinositol 4,5-Bisphosphate-containing Membranes and Inducing Cellular Transformation. J Biol Chem 287:5764\u20135774. https:\/\/doi.org\/10.1074\/jbc.M111.336487","journal-title":"J Biol Chem"},{"key":"10349_CR15","doi-asserted-by":"publisher","first-page":"W78","DOI":"10.1093\/nar\/gkae268","volume":"52","author":"I Letunic","year":"2024","unstructured":"Letunic I, Bork P (2024) Interactive Tree of Life (iTOL) v6: recent updates to the phylogenetic tree display and annotation tool. Nucleic Acids Res 52:W78\u2013W82. https:\/\/doi.org\/10.1093\/nar\/gkae268","journal-title":"Nucleic Acids Res"},{"key":"10349_CR16","doi-asserted-by":"publisher","first-page":"1123","DOI":"10.1126\/science.ade2574","volume":"379","author":"Z Lin","year":"2023","unstructured":"Lin Z, Akin H, Rao R et al (2023) Evolutionary-scale prediction of atomic-level protein structure with a language model. Science 379:1123\u20131130. https:\/\/doi.org\/10.1126\/science.ade2574","journal-title":"Science"},{"key":"10349_CR17","doi-asserted-by":"publisher","first-page":"1227","DOI":"10.1038\/s42003-025-08651-2","volume":"8","author":"GI Mart\u00ednez-Redondo","year":"2025","unstructured":"Mart\u00ednez-Redondo GI, Perez-Canales FM, Carbonetto B et al (2025) FANTASIA leverages language models to decode the functional dark proteome across the animal tree of life. Commun Biol 8:1227. https:\/\/doi.org\/10.1038\/s42003-025-08651-2","journal-title":"Commun Biol"},{"key":"10349_CR18","doi-asserted-by":"publisher","first-page":"1530","DOI":"10.1093\/molbev\/msaa015","volume":"37","author":"BQ Minh","year":"2020","unstructured":"Minh BQ, Schmidt HA, Chernomor O et al (2020) IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era. Mol Biol Evol 37:1530\u20131534. https:\/\/doi.org\/10.1093\/molbev\/msaa015","journal-title":"Mol Biol Evol"},{"key":"10349_CR19","doi-asserted-by":"publisher","first-page":"6379","DOI":"10.1016\/j.cell.2024.08.017","volume":"187","author":"J Morstein","year":"2024","unstructured":"Morstein J, Bowcut V, Fernando M et al (2024) Targeting Ras-, Rho-, and Rab-family GTPases via a conserved cryptic pocket. Cell 187:6379\u20136392e17. https:\/\/doi.org\/10.1016\/j.cell.2024.08.017","journal-title":"Cell"},{"key":"10349_CR20","doi-asserted-by":"publisher","first-page":"863","DOI":"10.1016\/S1097-2765(03)00356-3","volume":"12","author":"B Panic","year":"2003","unstructured":"Panic B, Perisic O, Veprintsev DB et al (2003) Structural Basis for Arl1-Dependent Targeting of Homodimeric GRIP Domains to the Golgi Apparatus. Mol Cell 12:863\u2013874. https:\/\/doi.org\/10.1016\/S1097-2765(03)00356-3","journal-title":"Mol Cell"},{"key":"10349_CR21","doi-asserted-by":"publisher","first-page":"giae055","DOI":"10.1093\/gigascience\/giae055","volume":"13","author":"C Pi\u00f1eiro","year":"2024","unstructured":"Pi\u00f1eiro C, Pichel JC (2024) Efficient phylogenetic tree inference for massive taxonomic datasets: harnessing the power of a server to analyze 1 million taxa. Gigascience 13:giae055. https:\/\/doi.org\/10.1093\/gigascience\/giae055","journal-title":"Gigascience"},{"key":"10349_CR22","doi-asserted-by":"publisher","first-page":"1995","DOI":"10.1073\/pnas.0908044107","volume":"107","author":"A Rausell","year":"2010","unstructured":"Rausell A, Juan D, Pazos F, Valencia A (2010) Protein interactions and ligand binding: From protein subfamilies to functional specificity. Proc Natl Acad Sci 107:1995\u20132000. https:\/\/doi.org\/10.1073\/pnas.0908044107","journal-title":"Proc Natl Acad Sci"},{"key":"10349_CR24","doi-asserted-by":"publisher","unstructured":"Rojas AM, Valencia A (2014) Evolution of the Ras superfamily of GTPases. Springer Vienna, pp 3\u201323. https:\/\/doi.org\/10.1007\/978-3-7091-1806-1_1","DOI":"10.1007\/978-3-7091-1806-1_1"},{"key":"10349_CR23","doi-asserted-by":"publisher","first-page":"189","DOI":"10.1083\/jcb.201103008","volume":"196","author":"AM Rojas","year":"2012","unstructured":"Rojas AM, Fuentes G, Rausell A, Valencia A (2012) The Ras protein superfamily: Evolutionary tree and role of conserved amino acids. J Cell Biol 196:189\u2013201. https:\/\/doi.org\/10.1083\/jcb.201103008","journal-title":"J Cell Biol"},{"key":"10349_CR25","doi-asserted-by":"publisher","first-page":"102577","DOI":"10.1016\/j.sbi.2023.102577","volume":"80","author":"L Santus","year":"2023","unstructured":"Santus L, Garriga E, Deorowicz S et al (2023) Towards the accurate alignment of over a million protein sequences: Current state of the art. Curr Opin Struct Biol 80:102577. https:\/\/doi.org\/10.1016\/j.sbi.2023.102577","journal-title":"Curr Opin Struct Biol"},{"issue":"6","key":"10349_CR26","doi-asserted-by":"publisher","first-page":"765","DOI":"10.1007\/s00239-025-10282-4","volume":"93","author":"R Shaw","year":"2025","unstructured":"Shaw R, Love SD, Mcwhite CD (2025) Evaluating Pretrained Protein Language Model Embeddings as Proxies for Functional Similarity. J Mol Evol 93(6):765\u2013776. https:\/\/doi.org\/10.1007\/s00239-025-10282-4","journal-title":"J Mol Evol"},{"key":"10349_CR27","doi-asserted-by":"publisher","unstructured":"Shi Y, Shen Y, Zhang X et al (2025) The ubiquitin code of RAS proteins: decoding its role in cancer progression. iScience 28. https:\/\/doi.org\/10.1016\/j.isci.2025.113029","DOI":"10.1016\/j.isci.2025.113029"},{"key":"10349_CR28","doi-asserted-by":"publisher","first-page":"761","DOI":"10.1042\/BCJ20250142","volume":"483","author":"DK Simanshu","year":"2026","unstructured":"Simanshu DK, McCormick F (2026) Revisiting RAS family GTPase signaling: effector selectivity and oncogenic bypass. Biochem J 483:761\u2013775. https:\/\/doi.org\/10.1042\/BCJ20250142","journal-title":"Biochem J"},{"key":"10349_CR29","doi-asserted-by":"publisher","first-page":"102997","DOI":"10.1016\/j.sbi.2025.102997","volume":"91","author":"K Weissenow","year":"2025","unstructured":"Weissenow K, Rost B (2025) Are protein language models the new universal key? Curr Opin Struct Biol 91:102997. https:\/\/doi.org\/10.1016\/j.sbi.2025.102997","journal-title":"Curr Opin Struct Biol"},{"key":"10349_CR30","doi-asserted-by":"publisher","first-page":"7","DOI":"10.1186\/1471-2105-15-7","volume":"15","author":"TJ Wheeler","year":"2014","unstructured":"Wheeler TJ, Clements J, Finn RD (2014) Skylign: a tool for creating informative, interactive logos representing sequence alignments and profile hidden Markov models. BMC Bioinformatics 15:7. https:\/\/doi.org\/10.1186\/1471-2105-15-7","journal-title":"BMC Bioinformatics"},{"key":"10349_CR31","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1124\/pr.117.014415","volume":"70","author":"C Yan","year":"2018","unstructured":"Yan C, Theodorescu D (2018) RAL GTPases: Biology and Potential as Therapeutic Targets in Cancer. Pharmacol Rev 70:1\u201311. https:\/\/doi.org\/10.1124\/pr.117.014415","journal-title":"Pharmacol Rev"}],"container-title":["Journal of Molecular Evolution"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00239-026-10349-w.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s00239-026-10349-w","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00239-026-10349-w.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,12]],"date-time":"2026-09-12T08:11:14Z","timestamp":1789200674000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s00239-026-10349-w"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,12]]},"references-count":31,"alternative-id":["10349"],"URL":"https:\/\/doi.org\/10.1007\/s00239-026-10349-w","relation":{},"ISSN":["0022-2844","1432-1432"],"issn-type":[{"value":"0022-2844","type":"print"},{"value":"1432-1432","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,12]]},"assertion":[{"value":"17 May 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"4 September 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"12 September 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors of this study do not report any conflicts of interest.","order":1,"name":"Ethics","label":"Conflict of Interest","group":{"name":"EthicsHeading","label":"Declarations"}}]}}