In silico biological discovery with large perturbation models.
| Title: | In silico biological discovery with large perturbation models. |
|---|---|
| Authors: | Miladinovic D; GSK plc, Zug, Switzerland. djordjemethz@gmail.com.; Höppe T; GSK plc, Zug, Switzerland.; Helmholtz Munich, Tübingen, Germany.; Chevalley M; GSK plc, Zug, Switzerland.; Georgiou A; GSK plc, Zug, Switzerland.; Stuart L; GSK plc, Zug, Switzerland.; Mehrjou A; GSK plc, Zug, Switzerland.; Bantscheff M; GSK plc, Zug, Switzerland.; Schölkopf B; Max Planck Institute for Intelligent Systems, Tübingen, Germany.; ELLIS Institute, Tübingen, Germany.; Schwab P; GSK plc, Zug, Switzerland. patrick.schwab@icloud.com. |
| Source: | Nature computational science [Nat Comput Sci] 2025 Nov; Vol. 5 (11), pp. 1029-1040. Date of Electronic Publication: 2025 Oct 15. |
| Publication Type: | Journal Article |
| Language: | English |
| Journal Info: | Publisher: Springer Nature Country of Publication: United States NLM ID: 101775476 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 2662-8457 (Electronic) Linking ISSN: 26628457 NLM ISO Abbreviation: Nat Comput Sci Subsets: MEDLINE |
| Imprint Name(s): | Original Publication: [New York, N.Y.] : Springer Nature, [2021]- |
| MeSH Terms: | Computational Biology*/methods ; Computer Simulation* ; Models, Biological*; Humans ; Gene Regulatory Networks ; Deep Learning ; Transcriptome |
| Abstract: | Data generated in perturbation experiments link perturbations to the changes they elicit and therefore contain information relevant to numerous biological discovery tasks-from understanding the relationships between biological entities to developing therapeutics. However, these data encompass diverse perturbations and readouts, and the complex dependence of experimental outcomes on their biological context makes it challenging to integrate insights across experiments. Here we present the large perturbation model (LPM), a deep-learning model that integrates multiple, heterogeneous perturbation experiments by representing perturbation, readout and context as disentangled dimensions. LPM outperforms existing methods across multiple biological discovery tasks, including in predicting post-perturbation transcriptomes of unseen experiments, identifying shared molecular mechanisms of action between chemical and genetic perturbations, and facilitating the inference of gene-gene interaction networks. LPM learns meaningful joint representations of perturbations, readouts and contexts, enables the study of biological relationships in silico and could considerably accelerate the derivation of insights from pooled perturbation experiments.; (© 2025. The Author(s).) |
| Competing Interests: | Competing interests: D.M., M.C., A.G., L.S., A.M., M.B. and P.S. are employees and shareholders of GSK plc. T.H. is a former employee of GSK plc. |
| References: | Can J Kidney Health Dis. 2016 Dec 14;3:2054358116679130. (PMID: 28781884); Nat Genet. 2025 Apr;57(4):797-808. (PMID: 40164735); Nat Biotechnol. 2024 Jun;42(6):927-935. (PMID: 37592036); Nat Rev Nephrol. 2019 Nov;15(11):713-726. (PMID: 31118499); Cell Syst. 2018 Dec 26;7(6):567-579.e6. (PMID: 30503647); Elife. 2021 Dec 07;10:. (PMID: 34874005); J Am Soc Nephrol. 2008 Sep;19(9):1659-62. (PMID: 18650476); Nucleic Acids Res. 2019 Jan 8;47(D1):D559-D563. (PMID: 30357367); Sci Signal. 2021 Feb 23;14(671):. (PMID: 33622983); Nat Methods. 2023 Nov;20(11):1759-1768. (PMID: 37770709); Cell. 2022 Jul 7;185(14):2559-2575.e28. (PMID: 35688146); Nat Commun. 2020 Aug 3;11(1):3877. (PMID: 32747659); Nucleic Acids Res. 2004 Jan 1;32(Database issue):D258-61. (PMID: 14681407); Clin Kidney J. 2018 Apr;11(2):254-258. (PMID: 29644067); Nat Rev Genet. 2015 May;16(5):299-311. (PMID: 25854182); Science. 2019 Aug 23;365(6455):786-793. (PMID: 31395745); Diabetes Care. 2004 Feb;27(2):468-73. (PMID: 14747230); Kidney Int. 2009 Apr;75(8):848-55. (PMID: 19165178); Cell. 2016 Dec 15;167(7):1853-1866.e17. (PMID: 27984732); Mol Syst Biol. 2023 Jun 12;19(6):e11517. (PMID: 37154091); Proc Natl Acad Sci U S A. 2016 Jul 5;113(27):7361-8. (PMID: 27382150); Nature. 1985 Oct 10-16;317(6037):542-4. (PMID: 2995836); Proc Natl Acad Sci U S A. 2007 Mar 13;104(11):4389-94. (PMID: 17360534); Cell. 2017 Nov 30;171(6):1437-1452.e17. (PMID: 29195078); Cell. 2018 Aug 9;174(4):953-967.e22. (PMID: 30033366); Nat Methods. 2023 Nov;20(11):1769-1779. (PMID: 37919419); Am J Epidemiol. 2016 Apr 15;183(8):758-64. (PMID: 26994063); BMC Genomics. 2019 Feb 4;20(Suppl 1):82. (PMID: 30712510); Commun Biol. 2025 Mar 11;8(1):412. (PMID: 40069299); J Am Soc Nephrol. 2008 Jan;19(1):102-8. (PMID: 18032793); Sci Rep. 2017 Mar 09;7:44219. (PMID: 28276528); Mol Syst Biol. 2020 Sep;16(9):e9198. (PMID: 32975352); Nat Biomed Eng. 2025 Apr;9(4):483-493. (PMID: 39643729); Cell Syst. 2021 Feb 17;12(2):128-140.e4. (PMID: 33373583); Clin J Am Soc Nephrol. 2021 May 8;16(5):790-799. (PMID: 32690722); Nucleic Acids Res. 2019 Jan 8;47(D1):D529-D541. (PMID: 30476227); Nucleic Acids Res. 2017 Jan 04;45(D1):D679-D686. (PMID: 27789686); Nature. 2023 Feb;614(7949):742-751. (PMID: 36755098); Cell. 2019 Jan 10;176(1-2):361-376.e17. (PMID: 30580963); Nucleic Acids Res. 2019 Jan 8;47(D1):D607-D613. (PMID: 30476243); Nat Methods. 2019 Aug;16(8):715-721. (PMID: 31363220); Nat Commun. 2023 Jan 24;14(1):390. (PMID: 36693830); Nat Methods. 2024 Aug;21(8):1470-1480. (PMID: 38409223); NPJ Syst Biol Appl. 2016;2:. (PMID: 28413689); Presse Med. 2011 Nov;40(11):1008-16. (PMID: 21981882); Nucleic Acids Res. 2018 Jan 4;46(D1):D649-D655. (PMID: 29145629); Cell. 2017 Jul 27;170(3):564-576.e16. (PMID: 28753430); Nature. 2023 Jun;618(7965):616-624. (PMID: 37258680); J Clin Invest. 2012 Nov;122(11):4257-73. (PMID: 23064367); Clin J Am Soc Nephrol. 2014 May;9(5):889-96. (PMID: 24721893); Kidney Res Clin Pract. 2022 Jul;41(4):422-431. (PMID: 35354242); Nat Rev Genet. 2010 Oct;11(10):733-9. (PMID: 20838408); Br J Anaesth. 2010 Feb;104(2):183-90. (PMID: 20086063); Nat Methods. 2020 Feb;17(2):147-154. (PMID: 31907445); J Clin Invest. 2015 Feb;125(2):607-20. (PMID: 25574838); J Am Soc Nephrol. 2022 Jun;33(6):1087-1104. (PMID: 35236775); Nat Methods. 2024 Aug;21(8):1481-1491. (PMID: 38844628); Nat Commun. 2018 Oct 17;9(1):4307. (PMID: 30333485); Transl Psychiatry. 2023 Apr 4;13(1):110. (PMID: 37015906); N Engl J Med. 2012 Dec 20;367(25):2407-18. (PMID: 23121377); Curr Control Trials Cardiovasc Med. 2000;1(3):161-165. (PMID: 11714433) |
| Molecular Sequence: | figshare GSE116198, GSE133344 |
| Entry Date(s): | Date Created: 20251015 Date Completed: 20251122 Latest Revision: 20260329 |
| Update Code: | 20260330 |
| PubMed Central ID: | PMC12638242 |
| DOI: | 10.1038/s43588-025-00870-1 |
| PMID: | 41094040 |
| Database: | MEDLINE |
Journal Article