Computationally Designed Nanobinders as Affinity Ligands in Diagnostic and Therapeutic Applications
@JCIM_JCTC
1. A groundbreaking study by Jeon et al. introduces the use of computationally designed nanobinders (DNBs) as a novel alternative to antibodies for detecting protein biomarkers in extracellular vesicles (EVs). This work leverages machine learning to design DNBs, which outperform antibodies in terms of signal intensity, sensitivity, and selectivity.
2. The study highlights the limitations of traditional antibodies, such as cross-reactivity and inconsistent affinity, which can lead to elevated background signals and undermine the accuracy of EV immunoassays. DNBs, with their unique design, minimize these issues.
3. The researchers developed a computational pipeline that includes RoseTTAFold diffusion for generating binder backbones, ProteinMPNN for converting structures into amino acid sequences, and AlphaFold-Multimer for predicting complex structures and evaluating confidence metrics.
4. The PD-L1-targeting DNB demonstrated superior performance in cellular imaging, with up to a 51-fold increase in signal intensity compared to antibodies. It also showed enhanced sensitivity and selectivity in EV analysis, making it a promising tool for cancer diagnostics.
5. Beyond diagnostics, the PD-L1 DNB exhibited therapeutic potential by effectively inhibiting immune checkpoints, suggesting its use in immune checkpoint blockade therapies.
6. The study’s findings underscore the potential of DNBs as a reliable and scalable platform for EV-based diagnostics and therapeutics, potentially advancing the field of cancer care and EV research.
📜Paper:
pubs.acs.org/doi/10.1021/jac…
#ComputationalBiology #Nanobinders #EVs #CancerDiagnostics #ImmuneCheckpointBlockade