Source 1primary paper2018PLoS Computational Biology
Toolkit/BROAD
BROAD
Taxonomy: Technique Branch / Method. Workflows sit above the mechanism and technique branches rather than replacing them.
Summary
BROAD is a computational protein design method that combines Rosetta-based structure modeling, machine learning, and integer linear programming to improve design search beyond Rosetta sampling alone. It was demonstrated in antibody design to increase the predicted HIV neutralization breadth of VRC23 across a panel of 180 divergent viral strains.
Usefulness & Problems
Why this is useful
BROAD is useful for protein design problems in which conventional Rosetta multistate design is limited by sampling and optimization across many target states. In the reported study, it increased predicted antibody breadth and reached 100% predicted binding across 180 HIV strains, indicating utility for broad-specificity design tasks.
Problem solved
BROAD was developed to overcome limitations in Rosetta sampling methods during multistate protein design. The demonstrated application addressed the challenge of increasing predicted HIV antibody neutralization breadth against diverse viral strains.
Taxonomy & Function
Primary hierarchy
Technique Branch
Method: A concrete computational method used to design, rank, or analyze an engineered system.
Mechanisms
integer linear optimizationmachine learning-guided designmultistate designstructure-based modelingTarget processes
recombinationImplementation Constraints
BROAD integrates structure-based modeling in the Rosetta software suite with machine learning and integer linear programming. The reported implementation was applied to multistate antibody design for VRC23 against a panel of divergent HIV strains, but the provided evidence does not specify model features, solver details, or required input formats.
The supplied evidence is limited to a single 2018 study and a computational antibody-design demonstration. The available evidence does not report experimental validation, implementation details beyond the algorithmic components, or performance in non-antibody protein systems.
Validation
Cell-freeBacteriaMammalianMouseHumanTherapeuticIndep. Replication
Supporting Sources
Ranked Claims
BROAD increased predicted breadth of antibody VRC23 against a panel of 180 divergent HIV viral strains and achieved 100% predicted binding against that panel.
We use this novel method to increase predicted breadth of naturally-occurring antibody VRC23 against a panel of 180 divergent HIV viral strains and achieve 100% predicted binding against the panel.
predicted binding coverage 100 %viral strain panel size 180 strains
BROAD increased predicted breadth of antibody VRC23 against a panel of 180 divergent HIV viral strains and achieved 100% predicted binding against that panel.
We use this novel method to increase predicted breadth of naturally-occurring antibody VRC23 against a panel of 180 divergent HIV viral strains and achieve 100% predicted binding against the panel.
predicted binding coverage 100 %viral strain panel size 180 strains
BROAD increased predicted breadth of antibody VRC23 against a panel of 180 divergent HIV viral strains and achieved 100% predicted binding against that panel.
We use this novel method to increase predicted breadth of naturally-occurring antibody VRC23 against a panel of 180 divergent HIV viral strains and achieve 100% predicted binding against the panel.
predicted binding coverage 100 %viral strain panel size 180 strains
BROAD increased predicted breadth of antibody VRC23 against a panel of 180 divergent HIV viral strains and achieved 100% predicted binding against that panel.
We use this novel method to increase predicted breadth of naturally-occurring antibody VRC23 against a panel of 180 divergent HIV viral strains and achieve 100% predicted binding against the panel.
predicted binding coverage 100 %viral strain panel size 180 strains
BROAD increased predicted breadth of antibody VRC23 against a panel of 180 divergent HIV viral strains and achieved 100% predicted binding against that panel.
We use this novel method to increase predicted breadth of naturally-occurring antibody VRC23 against a panel of 180 divergent HIV viral strains and achieve 100% predicted binding against the panel.
predicted binding coverage 100 %viral strain panel size 180 strains
BROAD increased predicted breadth of antibody VRC23 against a panel of 180 divergent HIV viral strains and achieved 100% predicted binding against that panel.
We use this novel method to increase predicted breadth of naturally-occurring antibody VRC23 against a panel of 180 divergent HIV viral strains and achieve 100% predicted binding against the panel.
predicted binding coverage 100 %viral strain panel size 180 strains
BROAD increased predicted breadth of antibody VRC23 against a panel of 180 divergent HIV viral strains and achieved 100% predicted binding against that panel.
We use this novel method to increase predicted breadth of naturally-occurring antibody VRC23 against a panel of 180 divergent HIV viral strains and achieve 100% predicted binding against the panel.
predicted binding coverage 100 %viral strain panel size 180 strains
BROAD significantly outperformed state-of-the-art multistate design in Rosetta in the reported comparison.
In addition, we compare the performance of this method to state-of-the-art multistate design in Rosetta and show that we can outperform the existing method significantly.
BROAD significantly outperformed state-of-the-art multistate design in Rosetta in the reported comparison.
In addition, we compare the performance of this method to state-of-the-art multistate design in Rosetta and show that we can outperform the existing method significantly.
BROAD significantly outperformed state-of-the-art multistate design in Rosetta in the reported comparison.
In addition, we compare the performance of this method to state-of-the-art multistate design in Rosetta and show that we can outperform the existing method significantly.
BROAD significantly outperformed state-of-the-art multistate design in Rosetta in the reported comparison.
In addition, we compare the performance of this method to state-of-the-art multistate design in Rosetta and show that we can outperform the existing method significantly.
BROAD significantly outperformed state-of-the-art multistate design in Rosetta in the reported comparison.
In addition, we compare the performance of this method to state-of-the-art multistate design in Rosetta and show that we can outperform the existing method significantly.
BROAD significantly outperformed state-of-the-art multistate design in Rosetta in the reported comparison.
In addition, we compare the performance of this method to state-of-the-art multistate design in Rosetta and show that we can outperform the existing method significantly.
BROAD significantly outperformed state-of-the-art multistate design in Rosetta in the reported comparison.
In addition, we compare the performance of this method to state-of-the-art multistate design in Rosetta and show that we can outperform the existing method significantly.
The authors state that BROAD is general and can be extended easily to other protein systems.
Finally, our approach is general and can be extended easily to other protein systems.
The authors state that BROAD is general and can be extended easily to other protein systems.
Finally, our approach is general and can be extended easily to other protein systems.
The authors state that BROAD is general and can be extended easily to other protein systems.
Finally, our approach is general and can be extended easily to other protein systems.
The authors state that BROAD is general and can be extended easily to other protein systems.
Finally, our approach is general and can be extended easily to other protein systems.
The authors state that BROAD is general and can be extended easily to other protein systems.
Finally, our approach is general and can be extended easily to other protein systems.
The authors state that BROAD is general and can be extended easily to other protein systems.
Finally, our approach is general and can be extended easily to other protein systems.
The authors state that BROAD is general and can be extended easily to other protein systems.
Finally, our approach is general and can be extended easily to other protein systems.
BROAD is a computational approach that combines Rosetta-based structural modeling with machine learning and integer linear programming to overcome limitations in Rosetta sampling methods.
We propose a new computational approach that combines traditional structure-based modeling using the Rosetta software suite with machine learning and integer linear programming to overcome limitations in the Rosetta sampling methods.
BROAD is a computational approach that combines Rosetta-based structural modeling with machine learning and integer linear programming to overcome limitations in Rosetta sampling methods.
We propose a new computational approach that combines traditional structure-based modeling using the Rosetta software suite with machine learning and integer linear programming to overcome limitations in the Rosetta sampling methods.
BROAD is a computational approach that combines Rosetta-based structural modeling with machine learning and integer linear programming to overcome limitations in Rosetta sampling methods.
We propose a new computational approach that combines traditional structure-based modeling using the Rosetta software suite with machine learning and integer linear programming to overcome limitations in the Rosetta sampling methods.
BROAD is a computational approach that combines Rosetta-based structural modeling with machine learning and integer linear programming to overcome limitations in Rosetta sampling methods.
We propose a new computational approach that combines traditional structure-based modeling using the Rosetta software suite with machine learning and integer linear programming to overcome limitations in the Rosetta sampling methods.
BROAD is a computational approach that combines Rosetta-based structural modeling with machine learning and integer linear programming to overcome limitations in Rosetta sampling methods.
We propose a new computational approach that combines traditional structure-based modeling using the Rosetta software suite with machine learning and integer linear programming to overcome limitations in the Rosetta sampling methods.
BROAD is a computational approach that combines Rosetta-based structural modeling with machine learning and integer linear programming to overcome limitations in Rosetta sampling methods.
We propose a new computational approach that combines traditional structure-based modeling using the Rosetta software suite with machine learning and integer linear programming to overcome limitations in the Rosetta sampling methods.
BROAD is a computational approach that combines Rosetta-based structural modeling with machine learning and integer linear programming to overcome limitations in Rosetta sampling methods.
We propose a new computational approach that combines traditional structure-based modeling using the Rosetta software suite with machine learning and integer linear programming to overcome limitations in the Rosetta sampling methods.
The modeled antibody variants were not tested in vitro, and the reported breadth increase is a prediction relative to wild-type antibody.
Although our modeled antibodies were not tested in vitro, we predict that these variants would have greatly increased breadth compared to the wild-type antibody.
The modeled antibody variants were not tested in vitro, and the reported breadth increase is a prediction relative to wild-type antibody.
Although our modeled antibodies were not tested in vitro, we predict that these variants would have greatly increased breadth compared to the wild-type antibody.
The modeled antibody variants were not tested in vitro, and the reported breadth increase is a prediction relative to wild-type antibody.
Although our modeled antibodies were not tested in vitro, we predict that these variants would have greatly increased breadth compared to the wild-type antibody.
The modeled antibody variants were not tested in vitro, and the reported breadth increase is a prediction relative to wild-type antibody.
Although our modeled antibodies were not tested in vitro, we predict that these variants would have greatly increased breadth compared to the wild-type antibody.
The modeled antibody variants were not tested in vitro, and the reported breadth increase is a prediction relative to wild-type antibody.
Although our modeled antibodies were not tested in vitro, we predict that these variants would have greatly increased breadth compared to the wild-type antibody.
The modeled antibody variants were not tested in vitro, and the reported breadth increase is a prediction relative to wild-type antibody.
Although our modeled antibodies were not tested in vitro, we predict that these variants would have greatly increased breadth compared to the wild-type antibody.
The modeled antibody variants were not tested in vitro, and the reported breadth increase is a prediction relative to wild-type antibody.
Although our modeled antibodies were not tested in vitro, we predict that these variants would have greatly increased breadth compared to the wild-type antibody.
Sequences recovered by BROAD recovered known binding motifs of broadly neutralizing anti-HIV antibodies.
We further demonstrate that sequences recovered by this method recover known binding motifs of broadly neutralizing anti-HIV antibodies.
Sequences recovered by BROAD recovered known binding motifs of broadly neutralizing anti-HIV antibodies.
We further demonstrate that sequences recovered by this method recover known binding motifs of broadly neutralizing anti-HIV antibodies.
Sequences recovered by BROAD recovered known binding motifs of broadly neutralizing anti-HIV antibodies.
We further demonstrate that sequences recovered by this method recover known binding motifs of broadly neutralizing anti-HIV antibodies.
Sequences recovered by BROAD recovered known binding motifs of broadly neutralizing anti-HIV antibodies.
We further demonstrate that sequences recovered by this method recover known binding motifs of broadly neutralizing anti-HIV antibodies.
Sequences recovered by BROAD recovered known binding motifs of broadly neutralizing anti-HIV antibodies.
We further demonstrate that sequences recovered by this method recover known binding motifs of broadly neutralizing anti-HIV antibodies.
Sequences recovered by BROAD recovered known binding motifs of broadly neutralizing anti-HIV antibodies.
We further demonstrate that sequences recovered by this method recover known binding motifs of broadly neutralizing anti-HIV antibodies.
Sequences recovered by BROAD recovered known binding motifs of broadly neutralizing anti-HIV antibodies.
We further demonstrate that sequences recovered by this method recover known binding motifs of broadly neutralizing anti-HIV antibodies.
Approval Evidence
1 source6 linked approval claimsfirst-pass slug broad
We propose a new computational approach that combines traditional structure-based modeling using the Rosetta software suite with machine learning and integer linear programming to overcome limitations in the Rosetta sampling methods. We demonstrate the effectiveness of this method, which we call BROAD
Source:
benchmark performancesupports
BROAD increased predicted breadth of antibody VRC23 against a panel of 180 divergent HIV viral strains and achieved 100% predicted binding against that panel.
We use this novel method to increase predicted breadth of naturally-occurring antibody VRC23 against a panel of 180 divergent HIV viral strains and achieve 100% predicted binding against the panel.
Source:
comparative performancesupports
BROAD significantly outperformed state-of-the-art multistate design in Rosetta in the reported comparison.
In addition, we compare the performance of this method to state-of-the-art multistate design in Rosetta and show that we can outperform the existing method significantly.
Source:
generalizabilitysupports
The authors state that BROAD is general and can be extended easily to other protein systems.
Finally, our approach is general and can be extended easily to other protein systems.
Source:
method descriptionsupports
BROAD is a computational approach that combines Rosetta-based structural modeling with machine learning and integer linear programming to overcome limitations in Rosetta sampling methods.
We propose a new computational approach that combines traditional structure-based modeling using the Rosetta software suite with machine learning and integer linear programming to overcome limitations in the Rosetta sampling methods.
Source:
predicted not experimentally testedsupports
The modeled antibody variants were not tested in vitro, and the reported breadth increase is a prediction relative to wild-type antibody.
Although our modeled antibodies were not tested in vitro, we predict that these variants would have greatly increased breadth compared to the wild-type antibody.
Source:
sequence feature recoverysupports
Sequences recovered by BROAD recovered known binding motifs of broadly neutralizing anti-HIV antibodies.
We further demonstrate that sequences recovered by this method recover known binding motifs of broadly neutralizing anti-HIV antibodies.
Source:
Comparisons
Source-backed strengths
The method significantly outperformed state-of-the-art multistate design in Rosetta in the reported comparison. In the benchmark application, BROAD increased the predicted breadth of antibody VRC23 and achieved 100% predicted binding against a 180-strain HIV panel.
Source:
In addition, we compare the performance of this method to state-of-the-art multistate design in Rosetta and show that we can outperform the existing method significantly.
Ranked Citations
- 1.