Back

A mechanism-aware stoichiometry platform resolves functional viral thresholds and induces antiviral hypersensitivity

Clinger, G.; Durden, H.; Daurelle, A.; Preece, B.; Peppel, W.; Gallegos, R.; Waheed, A. A.; Bohn, N.; MacArthur, B.; Petersen, A.; Yildiz, A.; Freed, E. O.; Saffarian, S.

2026-02-10 biophysics
10.64898/2026.02.09.704886 bioRxiv
Show abstract

Antiviral discovery is frequently hindered by a stoichiometric blind spot--a lack of resolution regarding the functional density of viral enzymes required for replication. Traditional screening pipelines rely on target-based or phenotypic assays that cannot distinguish between simple molecular binding and the crossing of a functional stoichiometric cliff. Here, we present a mechanism-aware platform that resolves these enzymatic requirements with single-virion precision. By integrating quantitative cryo-electron microscopy and genomic validation with Monte Carlo modeling, we map the stoichiometric landscapes of HIV-1 Protease (PR) and Reverse Transcriptase (RT). We uncover a striking disparity in enzymatic demand: a high-redundancy buffering capacity for PR ([~]40 monomers) contrasted with a high-threshold requirement for RT ([~]95 subunits). We demonstrate that by systematically de-buffering the virion, our platform induces a state of antiviral hypersensitivity, enabling the detection of therapeutic activity in novel and clinical inhibitors that remain invisible to traditional workflows. Furthermore, this multiplexed profiling enables de novo target identification, as inhibitors trigger failure exclusively on their respective stoichiometric arms. This platform provides a deterministic roadmap for de-risking drug discovery and identifying viral sub-stoichiometric vulnerabilities.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.