{"doi":"10.1038/s41586-024-07966-0","title":"The genetic architecture of protein stability","abstract":"<jats:title>Abstract</jats:title>\n                  <jats:p>\n                    There are more ways to synthesize a 100-amino acid (aa) protein (20\n                    <jats:sup>100</jats:sup>\n                    ) than there are atoms in the universe. Only a very small fraction of such a vast sequence space can ever be experimentally or computationally surveyed. Deep neural networks are increasingly being used to navigate high-dimensional sequence spaces\n                    <jats:sup>1</jats:sup>\n                    . However, these models are extremely complicated. Here, by experimentally sampling from sequence spaces larger than 10\n                    <jats:sup>10</jats:sup>\n                    , we show that the genetic architecture of at least some proteins is remarkably simple, allowing accurate genetic prediction in high-dimensional sequence spaces with fully interpretable energy models. These models capture the nonlinear relationships between free energies and phenotypes but otherwise consist of additive free energy changes with a small contribution from pairwise energetic couplings. These energetic couplings are sparse and associated with structural contacts and backbone proximity. Our results indicate that protein genetics is actually both rather simple and intelligible.\n                  </jats:p>","journal":"Nature","year":2024,"id":626194,"datarank":0.6190701577567639,"base_score":4.127134385045092,"endowment":4.127134385045092,"self_citation_contribution":0.6190701577567639,"citation_network_contribution":0.0,"self_endowment_contribution":0.6190701577567639,"citer_contribution":0.0,"corpus_percentile":null,"corpus_rank":null,"citation_count":61,"citer_count":0,"citers_with_citation_signal":0,"citers_with_endowment":0,"datacite_reuse_total":0,"is_dataset":false,"is_dataset_confidence":null,"is_data_producer":false,"deposit_databanks":null,"is_oa":false,"file_count":0,"downloads":0,"has_version_chain":false,"published_date":null,"fair_score":null,"fair_percentile":null,"algorithm_id":"datarank_citation_only_1hop_v6","ranking_scope":"data_only","authors":[{"id":1619727,"name":"Aina Martí-Aranda","orcid":null,"position":1,"is_corresponding":false},{"id":1619728,"name":"Cristina Hidalgo-Carcedo","orcid":"0000-0003-3571-2449","position":2,"is_corresponding":false},{"id":1526161,"name":"Antoni Beltran","orcid":"0000-0002-5949-2615","position":3,"is_corresponding":false},{"id":1619729,"name":"Jörn M. Schmiedel","orcid":"0000-0002-6842-9837","position":4,"is_corresponding":false},{"id":1328651,"name":"Ben Lehner","orcid":"0000-0002-8817-1124","position":5,"is_corresponding":false},{"id":1329558,"name":"André J. Faure","orcid":"0000-0002-4471-5994","position":0,"is_corresponding":false}],"reference_count":0,"raw_metadata":{"has_enrichment":true,"resolved":true,"title":"The genetic architecture of protein stability","abstract":"<jats:title>Abstract</jats:title>\n                  <jats:p>\n                    There are more ways to synthesize a 100-amino acid (aa) protein (20\n                    <jats:sup>100</jats:sup>\n                    ) than there are atoms in the universe. Only a very small fraction of such a vast sequence space can ever be experimentally or computationally surveyed. Deep neural networks are increasingly being used to navigate high-dimensional sequence spaces\n                    <jats:sup>1</jats:sup>\n                    . However, these models are extremely complicated. Here, by experimentally sampling from sequence spaces larger than 10\n                    <jats:sup>10</jats:sup>\n                    , we show that the genetic architecture of at least some proteins is remarkably simple, allowing accurate genetic prediction in high-dimensional sequence spaces with fully interpretable energy models. These models capture the nonlinear relationships between free energies and phenotypes but otherwise consist of additive free energy changes with a small contribution from pairwise energetic couplings. These energetic couplings are sparse and associated with structural contacts and backbone proximity. Our results indicate that protein genetics is actually both rather simple and intelligible.\n                  </jats:p>","is_dataset_classified":null,"base_score":4.110873864173311,"endowment":4.110873864173311,"datacite_reuse_total":0,"file_count":0,"downloads":0,"views":0,"has_version_chain":false,"is_dataset":false,"is_oa":false,"pmid":"39322666","pmcid":"PMC11499273","openalex_id":"https://openalex.org/W4402826776","authors":[],"funders":[{"funder_name":"European Research Council","grant_id":"883742","title":"Determining in vivo protein structures and understanding genetic interactions using deep mutagenesis"}],"total_grants":1,"fwci":9.5322,"citation_percentile":0.98941457,"influential_citations":0,"citation_trend":[{"year":2023,"count":1},{"year":2024,"count":5},{"year":2025,"count":34},{"year":2026,"count":20}],"oa_status":"hybrid","license":"cc-by","oa_locations":[{"url":"https://www.nature.com/articles/s41586-024-07966-0.pdf","host_type":"journal"},{"url":"https://www.nature.com/articles/s41586-024-07966-0.pdf","host_type":"publisher"},{"url":"https://www.nature.com/articles/s41586-024-07966-0","host_type":"publisher"},{"url":"https://doi.org/10.1038/s41586-024-07966-0","host_type":"journal"},{"url":"https://pubmed.ncbi.nlm.nih.gov/39322666","host_type":"repository"},{"url":"http://hdl.handle.net/10230/68884","host_type":"repository"},{"url":"https://www.ncbi.nlm.nih.gov/pmc/articles/11499273","host_type":"repository"},{"url":"https://repositori.upf.edu/bitstreams/f34d2602-da64-4f78-9c54-6654cf0b4577/download","host_type":"repository"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11499273/pdf/41586_2024_Article_7966.pdf","host_type":"repository"},{"url":"https://europepmc.org/articles/PMC11499273","host_type":"Europe_PMC"},{"url":"https://europepmc.org/articles/PMC11499273?pdf=render","host_type":"Europe_PMC"},{"url":"http://dx.doi.org/10.1038/s41586-024-07966-0","host_type":""}],"fields_of_study":["Protein Structure and Dynamics","RNA and protein synthesis mechanisms","Bioinformatics and Genomic Networks","0206 medical engineering","02 engineering and technology"],"mesh_terms":["Deep Learning","Computer Simulation","Humans","Models, Genetic","Phenotype","Proteins","Thermodynamics","Neural Networks, Computer","Protein Stability"],"keywords":["Pairwise comparison","Sequence (biology)","Simple (philosophy)","Sequence space","Space (punctuation)","Stability (learning theory)","Computer science","Sampling (signal processing)","Energy (signal processing)","Computational biology","Biological system","Physics","Biology","Artificial intelligence","Machine learning","Genetics","Mathematics","Pure mathematics","Models, Genetic","Protein Stability","Biophysics","Proteins","Genomics","Article","Computational biology and bioinformatics","Phenotype","Deep Learning","Thermodynamics","Humans","Computer Simulation","Neural Networks, Computer"],"sdg_mappings":[{"sdg_number":0,"sdg_label":"Affordable and clean energy"}],"linked_datasets":[],"clinical_trials":[],"software_tools":[],"database_accessions":[{"name":"pdb"}],"source":"live","citation_network_status":"fetched"},"created_at":"2026-08-04T12:50:15.656528Z","pmid":null,"pmcid":null,"fwci":null,"citation_percentile":null,"influential_citations":0,"oa_status":null,"license":null,"views":0,"total_file_size_bytes":0,"version_count":0,"fair_f":null,"fair_a":null,"fair_i":null,"fair_r":null,"fair_zscore":null,"fair_rationale":null,"fair_model":null,"fair_agent_version":null,"fair_fulltext_source":null,"fair_has_llm":null,"fair_computed_at":null,"clinical_trials":[],"software_tools":[],"db_accessions":[],"linked_datasets":[],"topics":[]}