Brian Koepnick

Brian Koepnick

Auto-generated profile, based on publicly available information

Brian Koepnick

Department of Biochemistry, University of Washington

Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 December 2024 in Science
Verified
Journal Article
Closed access
MGMatthias Glögl
Matthias Glögl
AKAditya Krishnakumar
Aditya Krishnakumar
RRRobert J. Ragotte
Robert J. Ragotte
IGInna Goreshnik
Inna Goreshnik
BCBrian Coventry
Brian Coventry
DBDavid Baker
David Baker
Despite progress in designing protein-binding proteins, the shape matching of designs to targets is lower than in many native protein complexes, and design efforts have failed for the tumor necrosis factor receptor 1 (TNFR1) and other protein targets with relatively flat and polar surfaces. We hypothesized that free diffusion from random noise could generate shape-matched binders for challenging targets and tested this approach on TNFR1. We obtain designs with low picomolar affinity whose specificity can be completely switched to other family members using partial diffusion. Designs function as antagonists or as superagonists when presented at higher valency for OX40 and 4-1BB. The ability to design high-affinity and high-specificity antagonists and agonists for pharmacologically important targets in silico presages a coming era in protein design in which binders are made by computation rather than immunization or random screening approaches.

4 June 2022 in bioRxiv
Verified
Preprint
Open access
PDF file
JDJ. Dauparas
J. Dauparas
IAI. Anishchenko
I. Anishchenko
NBN. Bennett
N. Bennett
HBH. Bai
H. Bai
RRR. J. Ragotte
R. J. Ragotte
DBD. Baker
D. Baker
While deep learning has revolutionized protein structure prediction, almost all experimentally characterized de novo protein designs have been generated using physically based approaches such as Rosetta. Here we describe a deep learning based protein sequence design method, ProteinMPNN, with outstanding performance in both in silico and experimental tests. The amino acid sequence at different positions can be coupled between single or multiple chains, enabling application to a wide range of current protein design challenges. On native protein backbones, ProteinMPNN has a sequence recovery of 52.4%, compared to 32.9% for Rosetta. Incorporation of noise during training improves sequence recovery on protein structure models, and produces sequences which more robustly encode their structures as assessed using structure prediction algorithms. We demonstrate the broad utility and high accuracy of ProteinMPNN using X-ray crystallography, cryoEM and functional studies by rescuing previously failed designs, made using Rosetta or AlphaFold, of protein monomers, cyclic homo-oligomers, tetrahedral nanoparticles, and target binding proteins. One-sentence summary: A deep learning based protein sequence design method is described that is widely applicable to current design challenges and shows outstanding performance in both in silico and experimental tests.

24 July 2020 in bioRxiv
Verified
Preprint
Open access
PDF file
CNChristoffer Norn
Christoffer Norn
BWBasile I. M. Wicky
Basile I. M. Wicky
DJDavid Juergens
David Juergens
SLSirui Liu
Sirui Liu
DKDavid Kim
David Kim
SOSergey Ovchinnikov
Sergey Ovchinnikov
The protein design problem is to identify an amino acid sequence which folds to a desired structure. Given Anfinsen’s thermodynamic hypothesis of folding, this can be recast as finding an amino acid sequence for which the lowest energy conformation is that structure. As this calculation involves not only all possible amino acid sequences but also all possible structures, most current approaches focus instead on the more tractable problem of finding the lowest energy amino acid sequence for the desired structure, often checking by protein structure prediction in a second step that the desired structure is indeed the lowest energy conformation for the designed sequence, and discarding the in many cases large fraction of designed sequences for which this is not the case. Here we show that by backpropagating gradients through the trRosetta structure prediction network from the desired structure to the input amino acid sequence, we can directly optimize over all possible amino acid sequences and all possible structures, and in one calculation explicitly design amino acid sequences predicted to fold into the desired structure and not any other. We find that trRosetta calculations, which consider the full conformational landscape, can be more effective than Rosetta single point energy estimations in predicting folding and stability of de novo designed proteins. We compare sequence design by landscape optimization to the standard fixed backbone sequence design methodology in Rosetta, and show that the results of the former, but not the latter, are sensitive to the presence of competing low-lying states. We show further that more funneled energy landscapes can be designed by combining the strengths of the two approaches: the low resolution trRosetta model serves to disfavor alternative states, and the high resolution Rosetta model, to create a deep energy minimum at the design target structure. Significance: Computational protein design has primarily focused on finding sequences which have very low energy in the target designed structure. However, what is most relevant during folding is not the absolute energy of the folded state, but the energy difference between the folded state and the lowest lying alternative states. We describe a deep learning approach which captures the entire folding landscape, and show that it can enhance current protein design methods.

12 November 2019 in PLOS Biology
Verified
Journal Article
Open access
PDF file
FKFiras Khatib
Firas Khatib
ADAmbroise Desfosses
Ambroise Desfosses
IKIstvan Kovanecz
Istvan Kovanecz
BKBrian Koepnick
Brian Koepnick
JFJeff Flatten
Jeff Flatten
Scott of HorowitzScott of Horowitz
Scott of Horowitz
With the rapid improvement of cryo-electron microscopy (cryo-EM) resolution, new computational tools are needed to assist and improve upon atomic model building and refinement options. This communication demonstrates that microscopists can now collaborate with the players of the computer game Foldit to generate high-quality de novo structural models. This development could greatly speed the generation of excellent cryo-EM structures when used in addition to current methods.

25 October 2016 in Nature Communications
Verified
Journal Article
Open access
PDF file
Scott of HorowitzScott of Horowitz
Scott of Horowitz
BKBrian Koepnick
Brian Koepnick
RMRaoul Martin
Raoul Martin
ATAgnes Tymieniecki
Agnes Tymieniecki
AWAmanda A. Winburn
Amanda A. Winburn
JBJames C. A. Bardwell
James C. A. Bardwell
Nature Communications 7: Article number:12549 (2016); Published 16 September 2016; Updated 25 October 2016 The original version of this Article contained an error in the spelling of a member of the University of Michigan students Consortium, Sam Lee, which was incorrectly given as Sam Lewe. This hasnow been corrected in both the PDF and HTML versions of the Article.

25 May 2022 in bioRxiv
Verified
Preprint
Open access
PDF file
EYErin C. Yang
Erin C. Yang
RDRobby Divine
Robby Divine
CKChristine S. Kang
Christine S. Kang
SCSidney Chan
Sidney Chan
EAElijah Arenas
Elijah Arenas
NKNeil P. King
Neil P. King
Undergraduate research experiences can improve student success in graduate education and STEM careers. During the COVID-19 pandemic, undergraduate researchers at our institution and many others lost their work-study research positions due to interruption of in-person research activities. This imposed a financial burden on the students and eliminated an important learning opportunity. To address these challenges, we created a paid, fully-remote, cohort-based research curriculum in computational protein design. Our curriculum used existing protein design methods as a platform to first educate and train undergraduate students and then to test research hypotheses. In the first phase, students learned computational methods to assess the stability of designed protein assemblies. In the second phase, students used a larger dataset to identify factors that could improve the accuracy of current protein design algorithms. This cohort-based program created valuable new research opportunities for undergraduates at our institute and enhanced the undergraduates’ feeling of connection with the lab. Students learned transferable and useful skills such as literature review, programming basics, data analysis, hypothesis testing, and scientific communication. Our program provides a model of structured computational research training opportunities for undergraduate researchers in any field for organizations looking to expand educational access. Graphical