Eva Smorodina
| Eva Smorodina | |
|---|---|
| Born | |
| 🏡 Residence | Oslo, Norway |
| 🎓 Alma mater | University of Oslo (PhD, 2026) |
| 💼 Occupation | |
| Known for | Computational structural biology; antibody engineering; membrane protein design |
Eva Smorodina is a computational immunologist and structural biologist based in Oslo, Norway. She defended her PhD thesis at the University of Oslo in August 2026. Her work looks at how antibodies bind their targets, how well AI models predict those interactions, and how membrane proteins can be redesigned to dissolve in water.[1]
Career
Smorodina joined Victor Greiff's group at the University of Oslo as a research assistant in July 2022 and started as a PhD student there in July 2024.[2] She defended her thesis, Structural rules of antibody function, on 28 August 2026, with Greiff as principal supervisor. The two opponents were Sarel-Jacob Fleishman of the Weizmann Institute of Science and Bruno Correia of EPFL. The university's summary says the thesis presents the first quantitative set of structural rules that separate high-affinity from low-affinity antibody binders. The rules still held when applied to modeled structures instead of experimental ones.[1]
Research
Antibody-antigen binding and structure prediction
In 2026 Smorodina was first author of a poster at an International Conference on Learning Representations (ICLR) workshop that asked whether AlphaFold3, Boltz-2 and Chai-1 can tell real antibody-antigen pairs from wrong ones. The models were run on 106 nanobody complexes with known structures and on more than 11,000 shuffled pairings. Their confidence scores often rated the wrong pairs as highly as the real ones, which produced many false positives.[3] The roughly 1.8 million predicted structures were deposited in the Sigma2 research data archive in April 2026.[4] A preprint describing this benchmark was discussed on the website of Science in a post headed "Predicting Antibody Binding: No Champagne Just Yet". The post concluded that computational prediction of antibody-antigen pairs is of little practical use for now.[5]
Earlier, she was one of 25 authors of Absolut!, a software suite that generates synthetic antibody-antigen binding structures for benchmarking machine learning methods. It appeared in Nature Computational Science in 2022.[6] Pieter Meysman wrote a News & Views piece about the paper in the same journal.[7]
Her other antibody papers include a 2022 review of machine-learning-based design of monoclonal antibodies in mAbs,[8] a 2024 analysis in Communications Biology of developability parameters across more than two million antibody sequences,[9] and AntBO, a Bayesian optimization method for designing antibody CDRH3 regions, in Cell Reports Methods.[10] In 2024 she also co-authored a benchmark of AlphaFold2-based workflows for predicting protein structural ensembles in Structure.[11]
Water-soluble membrane proteins
Smorodina worked with the group of Shuguang Zhang at the Massachusetts Institute of Technology on computational studies of the QTY code. The code swaps hydrophobic amino acids in transmembrane helices (leucine, isoleucine, valine and phenylalanine) for glutamine, threonine and tyrosine, which makes the proteins water soluble. She was first author of a 2022 study of 13 human solute carrier transporters in Scientific Reports[12] and of a 2024 study of human potassium channels in the same journal.[13] She is a co-author on related studies of bacterial outer membrane beta-barrel proteins,[14] ABC transporters[15] and glutamate transporters.[16]
She also co-authored a 2022 review in Chemical Reviews on designing proteins for water solubility and stability[17] and a 2023 Science Advances paper on a biosensor that pairs redesigned water-soluble membrane receptors with graphene transistor arrays.[18]
Other activities
In November 2024 the University of Oslo announced that Smorodina was one of 11 recipients of NOK 50,000 from its innovation funds for the idea and concept phase. Her project was titled "Supercharging antibodies with catalytic activity".[19]
She also took part in Nucleate Norway, a chapter of the biotech entrepreneurship network Nucleate, and helped organize its venture capital event in June 2025.[2] The chapter's website lists her as an alumna and as CEO of Yogutide.[20]
Awards
- 2024: Winner of the Antibody Society's Imaging Calendar Competition, for an image titled "Getting closer to binding with every step" that follows trastuzumab binding HER2 at increasing levels of detail.[21]
- 2024: Best poster prize at the BioCat conference in Stavanger, Norway.[2]
References
- ↑ 1.0 1.1 "Public defence: Eva Smorodina". Institute of Clinical Medicine, University of Oslo. Retrieved 30 September 2026.
- ↑ 2.0 2.1 2.2 "News". GreiffLab. Retrieved 30 September 2026.
- ↑ Smorodina E, Ali M, Kropivšek K, Salicari L, Miklavc S, Sormanni P, de Marco A, Greiff V (2026). Structural plausibility without binding specificity: limits of AI-based antibody-antigen structure prediction confidence scores. ICLR 2026 Workshop: Integrating Generative and Experimental Platforms for Biomolecular Design.
- ↑ Smorodina E, Kropivšek Brumat K, Ali M (16 April 2026). "Cognate and non-cognate antibody-antigen complexes: confidence scores". Sigma2 Archive. doi:10.11582/2026.owjnnp5q. Retrieved 30 September 2026.
- ↑ "Predicting Antibody Binding: No Champagne Just Yet". Science. American Association for the Advancement of Science. Retrieved 30 September 2026.
- ↑ Robert PA, Akbar R, Frank R, et al. (2022). "Unconstrained generation of synthetic antibody-antigen structures to guide machine learning methodology for antibody specificity prediction". Nature Computational Science. 2 (12): 845–865. doi:10.1038/s43588-022-00372-4. PMID 38177393 Check
|pmid=value (help). - ↑ Meysman, Pieter (19 December 2022). "Simulations that capture antigen-antibody complexity". Nature Computational Science. doi:10.1038/s43588-022-00379-x.
- ↑ Akbar R, Bashour H, Rawat P, et al. (2022). "Progress and challenges for the machine learning-based design of fit-for-purpose monoclonal antibodies". mAbs. 14 (1): 2008790. doi:10.1080/19420862.2021.2008790. PMID 35293269 Check
|pmid=value (help). - ↑ Bashour H, Smorodina E, Pariset M, et al. (2024). "Biophysical cartography of the native and human-engineered antibody landscapes quantifies the plasticity of antibody developability". Communications Biology. 7 (1): 922. doi:10.1038/s42003-024-06561-3. PMID 39085379 Check
|pmid=value (help). - ↑ Khan A, Cowen-Rivers AI, Grosnit A, et al. (2023). "Toward real-world automated antibody design with combinatorial Bayesian optimization". Cell Reports Methods. 3 (1): 100374. doi:10.1016/j.crmeth.2022.100374. PMID 36814835 Check
|pmid=value (help). - ↑ Riccabona JR, Spoendlin FC, Fischer AM, et al. (2024). "Assessing AF2's ability to predict structural ensembles of proteins". Structure. 32 (11): 2147–2159.e2. doi:10.1016/j.str.2024.09.001. PMID 39332396 Check
|pmid=value (help). - ↑ Smorodina E, Diankin I, Tao F, et al. (2022). "Structural informatic study of determined and AlphaFold2 predicted molecular structures of 13 human solute carrier transporters and their water-soluble QTY variants". Scientific Reports. 12 (1): 20103. doi:10.1038/s41598-022-23764-y. PMID 36418372 Check
|pmid=value (help). - ↑ Smorodina E, Tao F, Qing R, et al. (2024). "Computational engineering of water-soluble human potassium ion channels through QTY transformation". Scientific Reports. 14 (1): 28159. doi:10.1038/s41598-024-76603-7. PMID 39548172 Check
|pmid=value (help). - ↑ Sajeev-Sheeja A, Smorodina E, Zhang S (2023). "Structural bioinformatics studies of bacterial outer membrane beta-barrel transporters and their AlphaFold2 predicted water-soluble QTY variants". PLOS ONE. 18 (8): e0290360. doi:10.1371/journal.pone.0290360. PMID 37607179 Check
|pmid=value (help). - ↑ Pan E, Tao F, Smorodina E, Zhang S (2024). "Structural bioinformatics studies of six human ABC transporters and their AlphaFold2-predicted water-soluble QTY variants". QRB Discovery. 5: e1. doi:10.1017/qrd.2024.2. PMID 38577032 Check
|pmid=value (help). - ↑ Karagöl A, Karagöl T, Smorodina E, Zhang S (2024). "Structural bioinformatics studies of glutamate transporters and their AlphaFold2 predicted water-soluble QTY variants and uncovering the natural mutations of L->Q, I->T, F->Y and Q->L, T->I and Y->F". PLOS ONE. 19 (4): e0289644. doi:10.1371/journal.pone.0289644. PMID 38598436 Check
|pmid=value (help). - ↑ Qing R, Hao S, Smorodina E, Jin D, Zalevsky A, Zhang S (2022). "Protein design: from the aspect of water solubility and stability". Chemical Reviews. 122 (18): 14085–14179. doi:10.1021/acs.chemrev.1c00757. PMID 35921495 Check
|pmid=value (help). - ↑ Qing R, Xue M, Zhao J, et al. (2023). "Scalable biomimetic sensing system with membrane receptor dual-monolayer probe and graphene transistor arrays". Science Advances. 9 (29): eadf1402. doi:10.1126/sciadv.adf1402. PMID 37478177 Check
|pmid=value (help). - ↑ "Knowledge in use: NOK 4,5 million to innovative UiO researchers". University of Oslo. 15 November 2024. Retrieved 30 September 2026.
- ↑ "Norway (Associate)". Nucleate. Retrieved 30 September 2026.
- ↑ "Results of the Imaging Calendar Competition announced". The Antibody Society. 9 September 2024. Retrieved 30 September 2026.
Category:Living people
Category:Computational biologists
Category:Immunologists
Category:University of Oslo alumni
This article "Eva Smorodina" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:Eva Smorodina. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.
