{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,13]],"date-time":"2026-11-13T02:45:41Z","timestamp":1794537941715,"version":"4.9.1"},"reference-count":72,"publisher":"Elsevier BV","issue":"10","license":[{"start":{"date-parts":[[2025,10,1]],"date-time":"2025-10-01T00:00:00Z","timestamp":1759276800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/tdm\/userlicense\/1.0\/"},{"start":{"date-parts":[[2025,10,1]],"date-time":"2025-10-01T00:00:00Z","timestamp":1759276800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/legal\/tdmrep-license"},{"start":{"date-parts":[[2025,9,1]],"date-time":"2025-09-01T00:00:00Z","timestamp":1756684800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["cell.com","elsevier.com","sciencedirect.com"],"crossmark-restriction":true},"short-container-title":["iScience"],"published-print":{"date-parts":[[2025,10]]},"DOI":"10.1016\/j.isci.2025.113495","type":"journal-article","created":{"date-parts":[[2025,9,3]],"date-time":"2025-09-03T16:15:22Z","timestamp":1756916122000},"page":"113495","update-policy":"https:\/\/doi.org\/10.1016\/elsevier_cm_policy","source":"Crossref","is-referenced-by-count":6,"title":["Efficient inference, training, and fine-tuning of protein language models"],"prefix":"10.1016","volume":"28","author":[{"given":"Muhammed Hasan","family":"\u00c7elik","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5479-6345","authenticated-orcid":false,"given":"Xiaohui","family":"Xie","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"10.1016\/j.isci.2025.113495_bib1","doi-asserted-by":"crossref","first-page":"639","DOI":"10.1038\/s41580-024-00718-y","article-title":"Opportunities and challenges in design and optimization of protein function","volume":"25","author":"Listov","year":"2024","journal-title":"Nat. Rev. Mol. Cell Biol."},{"key":"10.1016\/j.isci.2025.113495_bib2","doi-asserted-by":"crossref","first-page":"1728","DOI":"10.1038\/s41467-022-29268-7","article-title":"Current progress and open challenges for applying deep learning across the biosciences","volume":"13","author":"Sapoval","year":"2022","journal-title":"Nat. Commun."},{"key":"10.1016\/j.isci.2025.113495_bib3","doi-asserted-by":"crossref","first-page":"583","DOI":"10.1038\/s41586-021-03819-2","article-title":"Highly accurate protein structure prediction with AlphaFold","volume":"596","author":"Jumper","year":"2021","journal-title":"Nature"},{"key":"10.1016\/j.isci.2025.113495_bib4","doi-asserted-by":"crossref","first-page":"706","DOI":"10.1038\/s41586-019-1923-7","article-title":"Improved protein structure prediction using potentials from deep learning","volume":"577","author":"Senior","year":"2020","journal-title":"Nature"},{"key":"10.1016\/j.isci.2025.113495_bib5","doi-asserted-by":"crossref","first-page":"493","DOI":"10.1038\/s41586-024-07487-w","article-title":"Accurate structure prediction of biomolecular interactions with AlphaFold 3","volume":"630","author":"Abramson","year":"2024","journal-title":"Nature"},{"key":"10.1016\/j.isci.2025.113495_bib6","doi-asserted-by":"crossref","first-page":"1123","DOI":"10.1126\/science.ade2574","article-title":"Evolutionary-scale prediction of atomic-level protein structure with a language model","volume":"379","author":"Lin","year":"2023","journal-title":"Science"},{"key":"10.1016\/j.isci.2025.113495_bib8","article-title":"ProtTrans: Towards Cracking the Language of Life\u2019s Code Through Self-Supervised Deep Learning and High Performance Computing","author":"Elnaggar","year":"2021","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib9","article-title":"Bilingual Language Model for Protein Sequence and Structure","author":"Heinzinger","year":"2024","journal-title":"bioRxiv"},{"key":"10.1016\/j.isci.2025.113495_bib10","doi-asserted-by":"crossref","first-page":"220","DOI":"10.1038\/s42256-024-00795-w","article-title":"Protein function prediction as approximate semantic entailment","volume":"6","author":"Kulmanov","year":"2024","journal-title":"Nat. Mach. Intell."},{"key":"10.1016\/j.isci.2025.113495_bib11","doi-asserted-by":"crossref","DOI":"10.1126\/science.adg7492","article-title":"Accurate proteome-wide missense variant effect prediction with AlphaMissense","volume":"381","author":"Cheng","year":"2023","journal-title":"Science"},{"key":"10.1016\/j.isci.2025.113495_bib12","doi-asserted-by":"crossref","first-page":"1512","DOI":"10.1038\/s41588-023-01465-0","article-title":"Genome-wide prediction of disease variant effects with a deep protein language model","volume":"55","author":"Brandes","year":"2023","journal-title":"Nat. Genet."},{"key":"10.1016\/j.isci.2025.113495_bib13","doi-asserted-by":"crossref","first-page":"1161","DOI":"10.1038\/s41588-018-0167-z","article-title":"Predicting the clinical impact of human mutation with deep neural networks","volume":"50","author":"Sundaram","year":"2018","journal-title":"Nat. Genet."},{"key":"10.1016\/j.isci.2025.113495_bib14","article-title":"Attention Is All You Need","author":"Vaswani","year":"2023","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib15","doi-asserted-by":"crossref","DOI":"10.7554\/eLife.82819","article-title":"Transformer-based deep learning for predicting protein properties in the life sciences","volume":"12","author":"Chandra","year":"2023","journal-title":"eLife"},{"key":"10.1016\/j.isci.2025.113495_bib16","article-title":"Language models enable zero-shot prediction of the effects of mutations on protein function","author":"Meier","year":"2021","journal-title":"bioRxiv"},{"key":"10.1016\/j.isci.2025.113495_bib17","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.2016239118","article-title":"Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences","volume":"118","author":"Rives","year":"2021","journal-title":"Proc. Natl. Acad. Sci.USA"},{"key":"10.1016\/j.isci.2025.113495_bib18","doi-asserted-by":"crossref","first-page":"968","DOI":"10.1016\/j.cels.2023.10.002","article-title":"ProGen2: Exploring the boundaries of protein language models","volume":"14","author":"Nijkamp","year":"2023","journal-title":"Cell Syst."},{"key":"10.1016\/j.isci.2025.113495_bib19","article-title":"MSA Transformer","author":"Rao","year":"2021","journal-title":"bioRxiv"},{"key":"10.1016\/j.isci.2025.113495_bib20","doi-asserted-by":"crossref","first-page":"2102","DOI":"10.1093\/bioinformatics\/btac020","article-title":"ProteinBERT: a universal deep-learning model of protein sequence and function","volume":"38","author":"Brandes","year":"2022","journal-title":"Bioinformatics"},{"key":"10.1016\/j.isci.2025.113495_bib21","article-title":"RoBERTa: A Robustly Optimized BERT Pretraining Approach","author":"Liu","year":"2019","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib22","article-title":"New mega dataset combined with deep neural network makes a progress in predicting impact of mutation on protein stability","author":"Pak","year":"2023","journal-title":"bioRxiv"},{"key":"10.1016\/j.isci.2025.113495_bib23","doi-asserted-by":"crossref","first-page":"2604","DOI":"10.1093\/molbev\/msz179","article-title":"GEMME: A Simple and Fast Global Epistatic Model Predicting Mutational Effects","volume":"36","author":"Laine","year":"2019","journal-title":"Mol. Biol. Evol."},{"key":"10.1016\/j.isci.2025.113495_bib24","doi-asserted-by":"crossref","first-page":"816","DOI":"10.1038\/s41592-018-0138-4","article-title":"Deep generative models of genetic variation capture the effects of mutations","volume":"15","author":"Riesselman","year":"2018","journal-title":"Nat. Methods"},{"key":"10.1016\/j.isci.2025.113495_bib25","doi-asserted-by":"crossref","first-page":"128","DOI":"10.1038\/nbt.3769","article-title":"Mutation effects predicted from sequence co-variation","volume":"35","author":"Hopf","year":"2017","journal-title":"Nat. Biotechnol."},{"key":"10.1016\/j.isci.2025.113495_bib26","doi-asserted-by":"crossref","first-page":"434","DOI":"10.1038\/s41586-023-06328-6","article-title":"Mega-scale experimental analysis of protein folding stability in biology and design","volume":"620","author":"Tsuboyama","year":"2023","journal-title":"Nature"},{"key":"10.1016\/j.isci.2025.113495_bib27","article-title":"PROSTATA: Protein Stability Assessment using Transformers","author":"Umerenkov","year":"2022","journal-title":"bioRxiv"},{"key":"10.1016\/j.isci.2025.113495_bib28","article-title":"FLIP: Benchmark tasks in fitness landscape inference for proteins","author":"Dallago","year":"2022","journal-title":"bioRxiv"},{"key":"10.1016\/j.isci.2025.113495_bib29","article-title":"SeqDance: A Protein Language Model for Representing Protein Dynamic Properties","author":"Hou","year":"2024","journal-title":"bioRxiv"},{"key":"10.1016\/j.isci.2025.113495_bib30","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.2320510121","article-title":"Leveraging a large language model to predict protein phase transition: A physical, multiscale, and interpretable approach","volume":"121","author":"Frank","year":"2024","journal-title":"Proc. Natl. Acad. Sci. USA"},{"key":"10.1016\/j.isci.2025.113495_bib31","doi-asserted-by":"crossref","first-page":"1358","DOI":"10.1126\/science.adf2465","article-title":"Enzyme function prediction using contrastive learning","volume":"379","author":"Yu","year":"2023","journal-title":"Science"},{"key":"10.1016\/j.isci.2025.113495_bib32","doi-asserted-by":"crossref","first-page":"8180","DOI":"10.1038\/s41467-024-52533-w","article-title":"Accurately predicting enzyme functions through geometric graph learning on ESMFold-predicted structures","volume":"15","author":"Song","year":"2024","journal-title":"Nat. Commun."},{"key":"10.1016\/j.isci.2025.113495_bib33","article-title":"Complex-based Ligand-Binding Proteins Redesign by Equivariant Diffusion-based Generative Models","author":"Nguyen","year":"2024","journal-title":"bioRxiv"},{"key":"10.1016\/j.isci.2025.113495_bib34","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.2316401121","article-title":"TULIP: A transformer-based unsupervised language model for interacting peptides and T cell receptors that generalizes to unseen epitopes","volume":"121","author":"Meynard-Piganeau","year":"2024","journal-title":"Proc. Natl. Acad. Sci. USA"},{"key":"10.1016\/j.isci.2025.113495_bib35","doi-asserted-by":"crossref","first-page":"7861","DOI":"10.1038\/s41467-023-43597-1","article-title":"ZeroBind: a protein-specific zero-shot predictor with subgraph matching for drug-target interactions","volume":"14","author":"Wang","year":"2023","journal-title":"Nat. Commun."},{"key":"10.1016\/j.isci.2025.113495_bib36","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.2220778120","article-title":"Contrastive learning in protein language space predicts interactions between drugs and protein targets","volume":"120","author":"Singh","year":"2023","journal-title":"Proc. Natl. Acad. Sci. USA"},{"key":"10.1016\/j.isci.2025.113495_bib37","article-title":"A unified framework for drug\u2013target interaction prediction by semantic-guided meta-path method","author":"Li","year":"2024","journal-title":"bioRxiv"},{"key":"10.1016\/j.isci.2025.113495_bib38","doi-asserted-by":"crossref","first-page":"275","DOI":"10.1038\/s41587-023-01763-2","article-title":"Efficient evolution of human antibodies from general protein language models","volume":"42","author":"Hie","year":"2024","journal-title":"Nat. Biotechnol."},{"key":"10.1016\/j.isci.2025.113495_bib39","article-title":"Simulating 500 million years of evolution with a language model","author":"Hayes","year":"2024","journal-title":"bioRxiv"},{"key":"10.1016\/j.isci.2025.113495_bib40","doi-asserted-by":"crossref","first-page":"91","DOI":"10.1038\/s41586-021-04043-8","article-title":"Disease variant prediction with deep generative models of evolutionary data","volume":"599","author":"Frazer","year":"2021","journal-title":"Nature"},{"key":"10.1016\/j.isci.2025.113495_bib41","article-title":"ScaleFold: Reducing AlphaFold Initial Training Time to 10 Hours","author":"Zhu","year":"2024","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib42","article-title":"Model Compression and Efficient Inference for Large Language Models: A Survey","author":"Wang","year":"2024","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib43","article-title":"Parameter-Efficient Fine-Tuning for Large Models: A Comprehensive Survey","author":"Han","year":"2024","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib44","article-title":"Efficient Large Language Models: A Survey","author":"Wan","year":"2024","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib45","article-title":"Beyond Efficiency: A Systematic Survey of Resource-Efficient Large Language Models","author":"Bai","year":"2024","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib46","article-title":"FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioning","author":"Dao","year":"2023","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib47","article-title":"ProteinGym: Large-Scale Benchmarks for Protein Design and Fitness Prediction","author":"Notin","year":"2023","journal-title":"bioRxiv"},{"key":"10.1016\/j.isci.2025.113495_bib48","doi-asserted-by":"crossref","first-page":"D523","DOI":"10.1093\/nar\/gkac1052","article-title":"UniProt: the Universal Protein Knowledgebase in 2023","volume":"51","year":"2023","journal-title":"Nucleic Acids Res."},{"key":"10.1016\/j.isci.2025.113495_bib49","article-title":"A Study of BFLOAT16 for Deep Learning Training","author":"Kalamkar","year":"2019","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib50","article-title":"Flex Attention: A Programming Model for Generating Optimized Attention Kernels","author":"Dong","year":"2024","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib51","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.2405840121","article-title":"Democratizing protein language models with parameter-efficient fine-tuning","volume":"121","author":"Sledzieski","year":"2024","journal-title":"Proc. Natl. Acad. Sci. USA"},{"key":"10.1016\/j.isci.2025.113495_bib52","doi-asserted-by":"crossref","first-page":"2880","DOI":"10.1038\/s41467-024-46947-9","article-title":"Genomic language model predicts protein co-regulation and function","volume":"15","author":"Hwang","year":"2024","journal-title":"Nat. Commun."},{"key":"10.1016\/j.isci.2025.113495_bib53","article-title":"LLM.int8: 8-bit Matrix Multiplication for Transformers at Scale","author":"Dettmers","year":"2022","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib54","article-title":"QLoRA: Efficient Finetuning of Quantized LLMs","author":"Dettmers","year":"2023","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib55","series-title":"bitsandbytes-foundation\/bitsandbytes: bitsandbytes","year":"2024"},{"key":"10.1016\/j.isci.2025.113495_bib56","article-title":"Pseudo-perplexity in One Fell Swoop for Protein Fitness Estimation","author":"Kantroo","year":"2024","journal-title":"bioRxiv"},{"key":"10.1016\/j.isci.2025.113495_bib57","article-title":"Convolutional Sequence to Sequence Learning","author":"Gehring","year":"2017","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib58","article-title":"Adam: A Method for Stochastic Optimization","author":"Kingma","year":"2014","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib59","article-title":"ZeRO-Offload: Democratizing Billion-Scale Model Training","author":"Ren","year":"2021","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib60","article-title":"LoRA: Low-Rank Adaptation of Large Language Models","author":"Hu","year":"2021","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib61","doi-asserted-by":"crossref","first-page":"495","DOI":"10.1038\/s41592-020-0801-4","article-title":"Meltome atlas\u2014thermal proteome stability across the tree of life","volume":"17","author":"Jarzab","year":"2020","journal-title":"Nat. Methods"},{"key":"10.1016\/j.isci.2025.113495_bib62","article-title":"DeepTFactor: A deep learning-based tool for the prediction of transcription factors","volume":"118","author":"Kim","year":"2021","journal-title":"Proc. Natl. Acad. Sci."},{"key":"10.1016\/j.isci.2025.113495_bib63","article-title":"ZeroQuant-V2: Exploring Post-training Quantization in LLMs from Comprehensive Study to Low Rank Compensation","author":"Yao","year":"2023","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib64","first-page":"27168","article-title":"ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformers","volume":"35","author":"Yao","year":"2022","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"10.1016\/j.isci.2025.113495_bib65","article-title":"FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness","author":"Dao","year":"2022","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib66","article-title":"PyTorch: An Imperative Style, High-Performance Deep Learning Library","author":"Paszke","year":"2019","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib67","doi-asserted-by":"crossref","first-page":"261","DOI":"10.1038\/s41592-019-0686-2","article-title":"SciPy 1.0: fundamental algorithms for scientific computing in Python","volume":"17","author":"Virtanen","year":"2020","journal-title":"Nat. Methods"},{"key":"10.1016\/j.isci.2025.113495_bib68","first-page":"2825","article-title":"Scikit-learn: Machine Learning in Python","volume":"12","author":"Pedregosa","year":"2011","journal-title":"J. Mach. Learn. Res."},{"key":"10.1016\/j.isci.2025.113495_bib69","article-title":"What Do Position Embeddings Learn? An Empirical Study of Pre-Trained Language Model Positional Encoding","author":"Wang","year":"2020","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib70","doi-asserted-by":"crossref","DOI":"10.1016\/j.neucom.2023.127063","article-title":"RoFormer: Enhanced Transformer with Rotary Position Embedding","volume":"568","author":"Su","year":"2024","journal-title":"Neurocomputing"},{"key":"10.1016\/j.isci.2025.113495_bib71","article-title":"Gaussian Error Linear Units (GELUs)","author":"Hendrycks","year":"2023","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib72","article-title":"GLU Variants Improve Transformer","author":"Shazeer","year":"2020","journal-title":"arXiv"},{"key":"10.1016\/j.isci.2025.113495_bib73","article-title":"On Layer Normalization in the Transformer Architecture","author":"Xiong","year":"2020","journal-title":"arXiv"}],"container-title":["iScience"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S2589004225017560?httpAccept=text\/xml","content-type":"text\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S2589004225017560?httpAccept=text\/plain","content-type":"text\/plain","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2026,2,20]],"date-time":"2026-02-20T17:38:18Z","timestamp":1771609098000},"score":1,"resource":{"primary":{"URL":"https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S2589004225017560"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10]]},"references-count":72,"journal-issue":{"issue":"10","published-print":{"date-parts":[[2025,10]]}},"alternative-id":["S2589004225017560"],"URL":"https:\/\/doi.org\/10.1016\/j.isci.2025.113495","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2024.10.22.619563","asserted-by":"object"}]},"ISSN":["2589-0042"],"issn-type":[{"value":"2589-0042","type":"print"}],"subject":[],"published":{"date-parts":[[2025,10]]},"assertion":[{"value":"Elsevier","name":"publisher","label":"This article is maintained by"},{"value":"Efficient inference, training, and fine-tuning of protein language models","name":"articletitle","label":"Article Title"},{"value":"iScience","name":"journaltitle","label":"Journal Title"},{"value":"https:\/\/doi.org\/10.1016\/j.isci.2025.113495","name":"articlelink","label":"CrossRef DOI link to publisher maintained version"},{"value":"article","name":"content_type","label":"Content Type"},{"value":"\u00a9 2025 The Author(s). Published by Elsevier Inc.","name":"copyright","label":"Copyright"}],"article-number":"113495"}}