Holberg, C. | Search bioRxiv

bioRxiv2026

What comes after de novo? Automated lead optimization of proteins with CRADLE-1

Lead optimization remains the longest and most expensive step in pre-clinical drug discovery, typically consuming 12-36 months whilst costing $5M-$15M per candidate. We introduce O_SCPLOWCRADLEC_SCPLOWO_SCPCAP-1C_SCPCAP, an automated framework for protein engineering. While O_SCPLOWCRADLEC_SCPLOWO_SCPCAP-1C_SCPCAP supports the full process of drug discovery and industrial protein engineering pipelines, including hit identification and de novo binder design, this work focuses on its application to multi-property lead optimization across protein modalities (VHHs, scFvs, IgGs, peptides, enzymes, CRISPR systems, vaccines). We show it is 4-7x faster than rational design, as measured by the number of wet lab rounds required. We provide in-vitro validation across all of the above modalities, typically optimizing multiple properties simultaneously (single and polyspecific binding down to picomolar, activity, thermostability,...). Technically, O_SCPLOWCRADLEC_SCPLOWO_SCPCAP-1C_SCPCAP starts with pre-trained foundation protein language models (PLMs), which are fine-tuned in unsupervised fashion on evolutionary neighborhoods, in supervised fashion using lab-in-the-loop data, and then deployed in a multi-model workflow. Of additional interest, we find that (a) the end-to-end system may be run in automated fashion; (b) wet lab data may be consumed in black box fashion without knowledge of the underlying biochemical mechanisms; (c) structural data may largely be superseded by sequence-function pairs.

bioengineering↗

bioRxiv2024

Cell-based high-content approach for SARS-CoV-2 neutralization identifies unique monoclonal antibodies and PI3K pathway inhibitors.

The sudden rise of the SARS-CoV-2 virus and the delay in the development of effective therapeutics to mitigate it made evident a need for ways to screen for compounds that can block infection and prevent further pathogenesis and spread. Yet, identifying effective drugs efficacious against viral infection and replication with minimal toxicity for the patient can be difficult. Monoclonal antibodies were shown to be effective, yet as the SARS-CoV-2 mutated, these antibodies became ineffective. Small molecule antivirals were identified using pseudovirus constructs to recapitulate infection in non-human cells, such as Vero E6 cells. However, the impact was limited due to poor translation of these compounds in the clinical setting. This is partly due to the lack of similarity of screening platforms to the in vivo physiology of the patient and partly because drugs effective in vitro showed dose-limiting toxicities. In this study, we performed two high-throughput screens in human lung adenocarcinoma cells with authentic SARS-CoV-2 virus to identify both monoclonal antibodies that neutralize the virus and clinically useful kinase inhibitors to block the virus and prioritize minimal host toxicity. Using high-content imaging combined with single-cell and multidimensional analysis, we identified antibodies and kinase inhibitors that reduce virus infection without affecting the host. Our screening technique uncovered novel antibodies and overlooked kinase inhibitors (i.e. PIK3i, mTORi, multiple RTKi) that could be effective against SARS-CoV-2 virus. Further characterization of these molecules will streamline the repurposing of compounds for the treatment of future pandemics and uncover novel mechanisms viruses use to hijack and infect host cells.

pharmacology and toxicology↗